# My Sick Builds

My Sick Builds is a boutique AI and automation consultancy based in the Nashville area, run by Cory Micek. MSB builds three core systems for service businesses: AI Operating Systems (custom dashboards and orchestration layers that give businesses an AI-powered command center), Workflow Systems (n8n and Trigger.dev automations that replace manual processes with reliable, measurable pipelines), and Voice Systems (AI voice agents for inbound and outbound client communication). The practice focuses on regulated and relationship-driven verticals including HVAC, roofing, insurance, law firms, and driving schools.

## My Sick Builds: AI Operating Systems for Service Businesses
URL: https://mysickbuilds.com/
Description: AI Operating Systems, Workflow Systems, and Voice Systems for service businesses.

### Services

- **AI Operating Systems (AIOS)**: Custom dashboards and orchestration layers that give a business an AI-powered command center for its tools, data, and agents.
- **Workflow Automation**: n8n and Trigger.dev automations that replace brittle manual processes with reliable, measurable pipelines.
- **AI Voice Agents**: Human-sounding AI voice agents that answer calls, qualify leads, book jobs, and follow up for inbound and outbound communication.

### Frequently asked questions

**Q: What can you build?**
A: If your team does it by hand every week, we can probably build a system for it. The three we build most: AI Operating Systems (a team of AI agents and one dashboard to run the business), Workflow Automations that kill repetitive data entry and Voice Agents that answer the phone, qualify leads and book appointments while you sleep. The audit tells us which one pays for itself first.

**Q: Will it integrate with my existing tools?**
A: That is the point. We build on top of the tools you already use (CRM, scheduling, POS, phones and more) instead of forcing you onto something new. We only suggest a new tool when it is genuinely faster, simpler, or cheaper for you.

**Q: How long until something is live?**
A: First working build is typically weeks, not months. The audit comes first so we sequence the highest-ROI piece to ship early. The goal is to land quick wins with you up front, then build from there.

**Q: What does this cost?**
A: Pricing is flat and predictable, not hourly, with no surprise fees. The audit gives you the exact scope and numbers before you commit to anything.

**Q: Do I own it, or am I locked in?**
A: You choose. Two paths: a lower monthly subscription that covers the build, hosting and ongoing changes, or a flat build rate where you own the system outright with no recurring fee to us. We walk you through the tradeoffs in the audit and you pick what fits your budget.

**Q: Is my data secure?**
A: Yes. We give each system only the access it actually needs, keep your data in accounts you own and control whenever possible and keep a clear record of what the system does so nothing is a black box.

**Q: What happens when the AI does not know the answer?**
A: It escalates gracefully, to a human, a callback, or a fallback flow you define. No dead ends, no invented promises. From there we train the AI so it does not happen again.

**Q: Do I need technical staff to run it?**
A: No. We operate and monitor the systems and you get a dashboard plus a human (Cory) on the other end. We are happy to train your team to run it. If you have technical staff, we will walk them through exactly how it works. Most systems run on their own, but when you need additions or changes, our maintenance option has you covered.

**Q: What if I only want one piece, not the whole system?**
A: Totally fine. Many clients start with their biggest problem, a single system or workflow and expand once they see the return.

**Q: What is the difference between the Audit and the AI Opportunity Map?**
A: Two rungs on the same ladder. The AI Opportunity Map you can run yourself: answer a few questions, get back a ranked list of exactly where AI can buy back hours and dollars in your business. The Audit is the step up, a one-on-one with you and your team where we take that map and turn it into a sequenced build plan with numbers attached. The Map shows you what is possible. The Audit shows you what to build first and what it is worth. Natural move: get your Map, then book a call to go through it with me.

**Q: Why should I work with you?**
A: Most people selling AI right now are winging it. You work directly with me, Cory. I spent 25 years building apps and dashboards for Amazon, Marriott, GE, Fidelity and many other Fortune 500 companies, the systems that quietly power the world. So you're not getting someone who watched a few videos and bought a course. You're getting an operator who has shipped real systems at scale, sitting at your table, building for your business. The person who scopes your build is the person who builds it.

## Blog
URL: https://mysickbuilds.com/blog/
Description: Notes from the build floor: practical AI, automation, and systems thinking for service businesses. Written between builds, not from an ivory tower.

## Case Studies
URL: https://mysickbuilds.com/case-studies/
Description: Real builds. Real receipts. Every system we ship gets a case file: what it replaced, what it costs to run, and what it gives back.

## Free AI Audit
URL: https://mysickbuilds.com/ai-opportunity-map/
Description: A free 5-minute AI audit for your business: where you lose hours, what to automate first, and what it's worth. Answer 19 quick questions and Cory reads every one personally.

## AI Operating System for Your Business
URL: https://mysickbuilds.com/ai-operating-system/
Description: One monthly number gets you an AI department: a dashboard, agents your team can talk to, and automations that answer leads in under a minute. Pricing included.

### Frequently asked questions

**Q: What is an AI Operating System (AIOS)?**
A: An AI Operating System is an AI department installed into your business: a command-center dashboard, AI agents your team can talk to, and automations that handle lead response, follow-up, reviews, and reporting. You pay one monthly number and My Sick Builds owns and manages all of the technology.

**Q: How much does an AI Operating System cost?**
A: Three subscription tiers: Starter at $2,500/month with 3 workflows installed, Pro at $5,000/month with 7 workflows, and Complete at $10,500/month with 15 or more. Every engagement starts with a $3,500 paid discovery that is credited back in full across your first two months.

**Q: What does the $3,500 discovery include?**
A: An AI Opportunity Map of your business, a demo dashboard mocked up with your own use cases, a workflow selection from the catalog, and a signed-off onboarding plan. If you subscribe, the full $3,500 is credited across months 1 and 2, so the discovery pays for itself the moment you say yes.

**Q: Is there a contract or minimum term?**
A: Yes: a 6-month minimum, then month-to-month with 30 days notice. Month 1 is due at signing. Every tier includes changes to your installed workflows, one catalog swap per quarter, monthly maintenance, and your client-branded dashboard.

**Q: Do I ever own the system?**
A: There is a buyout path, and the number goes in your contract so you are never locked in. The infrastructure runs in accounts you can see, and nothing is hidden inside a bundle.

**Q: Who is an AIOS not for?**
A: Businesses under about 5 people with simple operations usually need a single automation, not a system. If that is you, a one-off workflow or voice agent build is the honest recommendation, and it costs a fraction of a subscription.

## AIOS Workflow Library
URL: https://mysickbuilds.com/workflow-library/
Description: The workflow library behind the AI Operating System: speed-to-lead, missed-call textback, review engines, AI intake, and 50+ more, installed by tier and swappable quarterly.

### Frequently asked questions

**Q: What is the AIOS workflow library?**
A: It's the catalog of 50+ pre-built automations that AI Operating System subscriptions install from: speed-to-lead response, missed-call textback, review engines, AI intake, invoice chasing, and vertical-specific workflows for home services, law firms, and clinics. Starter installs 3, Pro installs 7, Complete installs 15 or more.

**Q: Can I swap workflows after they're installed?**
A: Yes. Every AIOS tier includes one catalog swap per quarter plus ongoing changes to your installed workflows, so the system keeps matching your business instead of freezing at day one.

**Q: What if the workflow I need isn't in the library?**
A: I build it custom. You name the job, I build the agent that does it, and it becomes part of your system. The library itself grew out of exactly those requests.

**Q: How do I know which workflows to start with?**
A: That's what the paid discovery is for: it maps where your hours and leads are actually leaking and selects your first workflows from the catalog. Your first three always include at least one directly tied to revenue.

## About Cory Micek
URL: https://mysickbuilds.com/about/
Description: Cory Micek spent 25 years building for Amazon, Marriott, and GE. Now he brings enterprise-grade AI to small businesses, without the enterprise price tag.

### Frequently asked questions

**Q: Who is Cory Micek?**
A: Cory Micek is the founder and AI Solutions Architect of My Sick Builds, an AI consultancy in Nashville, TN. He spent 25 years building technology for enterprise companies including Amazon, Marriott, GE, Johnson & Johnson, Fidelity, and Verizon before turning that experience toward small businesses.

**Q: What does My Sick Builds do?**
A: My Sick Builds designs and builds AI Operating Systems, workflow automation, and AI voice agents for service businesses. Cory audits where a business is losing time, maps a custom AI roadmap, and builds the systems himself on flat pricing you can own or subscribe to.

**Q: Why hire a solo builder instead of an agency?**
A: The person who scopes your system is the same person who builds and supports it. One architect, start to finish: no handoffs, no junior teams, and 25 years of enterprise engineering standards applied directly to your build.

**Q: How do I get started with My Sick Builds?**
A: Book a free discovery call. It is 15 minutes with Cory to walk through where your week actually goes, and you leave with clear next steps whether you build together or not.

## Contact
URL: https://mysickbuilds.com/contact/
Description: Tell us what you're trying to fix. We reply within 24 hours.

## Book a Discovery Call
URL: https://mysickbuilds.com/book/
Description: 15 minutes with Cory. Walk through where your week actually goes and leave with next steps, whether we build or not.

## How Do I Know If My Business Needs an AI Voice Agent?
URL: https://mysickbuilds.com/blog/does-my-business-need-an-ai-voice-agent/
Date: August 4, 2026

Here's the fastest honest answer: if your phone rings more than someone can answer it, or it rings when nobody's there to answer at all, you have the problem a voice agent solves. If neither is true, you probably don't, and you can stop reading in two paragraphs. The rest of this post is the self-assessment I'd run on your business in a discovery call, including the signals that say *don't* buy one.

- **~80%** of voicemail callers leave no message
- **168 hrs** a week your phone can ring
- **4** signals that say you need one

## What are the signals you need a voice agent?

Four show up over and over in the businesses I build for:

**Calls ring out when you're busiest.** The cruel math of a service business: the weeks with the most calls are the weeks you're least able to answer them. Every ring-out during a heat wave or a freeze is your highest-intent caller of the year, and [roughly 80% of callers who hit voicemail don't leave a message](https://www.numa.com/blog/22-business-phone-statistics). They don't call back later; they call the next name in the search results.

**Real demand arrives after hours.** A dead water heater at 11pm, a fender bender on Saturday, a toothache on Sunday night. If your industry has emergencies or your customers book outside 9-to-5, coverage is the product, and a human schedule can't provide it.

**Your intake is the same script every call.** Name, address, symptom, urgency, insurance, callback window. If you could write your phone intake down as a numbered list, it's automatable, and doing it by hand is costing someone hours a week.

**One person IS the phone system.** Often the owner. If stepping into a job site, a meeting, or a vacation means the business stops answering, you don't have a coverage plan, you have a single point of failure with a personal cell.

Score yourself honestly:

## When do you NOT need one?

When coverage is already solved. A handful of calls a week that a human reliably answers is not a problem worth a build, and I tell prospects that in the first fifteen minutes. Two other honest disqualifiers:

| Situation | Why a voice agent won't help yet | Do this instead |
|---|---|---|
| **Low volume, solved coverage** | The build never pays back on ten calls a week | Keep your setup; revisit at real volume |
| **No defined intake process** | An agent automates your script; no script, nothing to automate | Write the intake down first, then wire it |
| **The problem is upstream** | If the phone isn't ringing, answering it better fixes nothing | Marketing and follow-up first: [does automation pay?](/blog/can-workflow-automation-save-money/) |

An agent is a coverage tool, not a growth hack. It converts demand you're already losing; it doesn't create demand you don't have.

## What does "yes" actually look like?

The clearest case I've built: a plumbing company whose owner answered the same 13-step intake himself, call after call, including the 2am emergencies that are the highest-margin jobs a plumber gets. The voice agent runs that exact intake on every call now, books the job, and posts a summary for a quick human audit, and it [gave him back more than 20 hours a week](/case-studies/instant-water-heaters/). Nothing about his demand changed. The only thing that changed is that answering stopped depending on a human being awake.

If you scored two or more on the checklist, here's the path from yes to live:

1. **Run the missed-call math** Calls missed last month times the value of one booked job. That number is your monthly leak, and it's the budget conversation.
2. **Map the one call flow that bleeds most** Usually after-hours or overflow. One flow, scripted end to end, not your whole phone system.
3. **Build, test on real calls, wire the trust loop** The agent books into your calendar and CRM, and every booking posts for a quick human check. You inspect its work; you don't take it on faith.
4. **Expand only after it's proven** Once the first flow pays, add overflow, outbound reminders, or the daytime line. Each step earns the next.

The costs are covered honestly, line by line, in [what an AI voice agent actually costs](/blog/how-much-does-an-ai-voice-agent-cost/), and if you're wondering whether the booking part really works without a human, [that has its own post too](/blog/can-ai-voice-agents-book-appointments/).

Still not sure which side of the checklist you're on? [Book the 15-minute call](/book/) and bring last month's call log. I'll run the math with you, and if the honest answer is "your receptionist has it handled," you'll hear exactly that.

## What Types of Workflows Should You Automate First?
URL: https://mysickbuilds.com/blog/what-workflows-should-i-automate-first/
Date: July 30, 2026

The best first automation is almost never the flashiest one. It's the chore you'd be embarrassed to admit how much time you spend on: the invoice chasing, the same reminder text, the lead that sat until Thursday. [McKinsey's automation research](https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/four-fundamentals-of-workplace-automation) found that in about 60% of jobs, at least a third of the tasks could already be automated with existing technology. This post is the filter for finding which third of *yours* goes first. If you're still deciding whether automation pays at all, [start with the ROI math](/blog/can-workflow-automation-save-money/) and come back.

- **104 hrs/yr** one farm recovered from ONE workflow
- **60%** of jobs: a third of tasks automatable
- **1** the right number of workflows to start with

## What makes a workflow worth automating first?

Two questions decide it: how often does the task run, and how much judgment does it need? Plot every chore in your week on those two axes and the answer stops being a debate:

The top-right quadrant, repeated and predictable, is the only place a *first* automation belongs. Those tasks have written-downable steps and run weekly or daily, so the build pays for itself on volume. The judgment-heavy quadrants aren't off-limits forever; they just come later, with a human approval step wired in, once the easy wins have paid for the plumbing.

## Which five workflows pay first for most businesses?

Across the small businesses I work with (home services, farms, professional practices), the same five candidates keep winning. The tell for each one is a symptom you already recognize:

| Workflow | The symptom you have it | What the automation does |
|---|---|---|
| **Lead follow-up** | Inquiries answered "when I get a minute," which is tomorrow | Instant reply + booking link the moment a lead lands |
| **Appointment reminders** | No-shows, and mornings spent confirming by hand | Automatic text/email sequence before every appointment |
| **Invoicing + payment chasing** | You are the accounts-receivable department, at 9pm | Invoices generated and politely chased on a schedule |
| **Review requests** | Happy customers, thin Google profile | A well-timed ask after every completed job |
| **Recurring reports** | Copy-paste from three systems into one spreadsheet | The numbers assemble themselves and arrive by email |

If several apply, pick the one closest to revenue. A no-show costs you a slot; a lead that goes cold costs you the whole job.

## How do you actually find yours?

Don't guess from a list, audit a week. The procedure is short:

1. **Track one honest week** Every time a task repeats, write it down with a minutes count. No fixing yet, just counting.
2. **Filter with the 2x2** Keep only the repeated, predictable ones. If you couldn't hand it to a new hire with written steps, it's not first.
3. **Rank by bleed** Multiply minutes by frequency, and weight anything that touches leads or cash flow. The top of that list is your first automation.
4. **Fix the process on paper** Clean up the steps BEFORE wiring anything. Automating a broken process just makes mistakes faster.
5. **Build one, measure, expand** Prove it with a before-and-after number, then let workflow two reuse the same plumbing.

That fourth step is the one everyone skips. Automation is an amplifier: point it at a clean process and it multiplies the output, point it at a mess and it multiplies the mess.

> **The honest filter.** If you can't write the steps down clearly enough for a new hire to follow them, it's not ready to automate. That's not a software problem, and no tool fixes it.

## What does a good first automation look like in practice?

A cattle farm I built for had a classic top-right-quadrant chore: weekly inventory meant 90 minutes in a freezer reading box weights to a helper, then another hour of someone retyping it all into spreadsheets. Repeated weekly, completely predictable, three people involved. We replaced it with [one phone call to a voice agent that writes everything to Google Sheets and Square directly](/case-studies/mooberry-farms/): 2.5 hours became 31 minutes, on the first run, and the owner got roughly 104 hours a year back from that single workflow.

Notice what made it the *right* first pick: it ran every week, the steps never changed, and nothing about it needed judgment. Weighing a box is not a decision. That's the pattern to hunt for in your own week.

Starting with one workflow isn't just less risky, it's cheaper: a single-purpose automation is [a small fraction of what a full multi-agent system costs](/blog/how-much-does-an-ai-operating-system-cost/), and the accounts and integrations it sets up get reused by everything you add later.

If you want the audit done for you, [book the 15-minute call](/book/) and walk me through your week. I'll tell you which workflow I'd automate first and what it would cost, and if the honest answer is "clean up the process and spend nothing yet," that's the answer you'll get.

## How Much Does It Cost to Run an AI Operating System?
URL: https://mysickbuilds.com/blog/how-much-does-an-ai-operating-system-cost/
Date: July 28, 2026

The content engine inside one of my AI operating system builds writes a researched, SEO-ready blog post for about six cents in AI fees. An agency charges $150 to $500 for the same post. That one line tells you most of what this article explains: running an AI operating system is cheap, and the real money is in building it. Here are the actual numbers for both, and the honest way to think about whether the build is worth it. If you're still working out what an AIOS even is, [start with the plain-English explainer](/blog/what-is-an-ai-operating-system/) and come back.

- **$60-100/yr** per cloud server (a build uses ~2)
- **~$0.06** AI fees per SEO blog post
- **pennies** per task, metered like a utility

## What does an AI operating system cost to run each month?

Less than most owners' software subscriptions. The recurring costs are cloud hosting, metered AI model fees, and a handful of small third-party tools, and every one of them is itemized in your name. Here's the actual line-item list from my builds:

| Line item | What it is | What it runs |
|---|---|---|
| **Cloud servers** | One for your website, one where the agents live | [$60 to $100 a year each](https://www.hostinger.com/vps-hosting) |
| **AI model fees** | Metered per task, like a utility | Pennies per task; [priced per million tokens](https://claude.com/pricing) |
| **Third-party tools** | The connectors a workflow touches | A few dollars a month per workflow |
| **Voice infrastructure** | Phone numbers, telephony, minutes (voice builds only) | $150 to $1,200+ a month, usage-driven |
| **Maintenance plan** | Optional: monitoring, updates, improvements | Optional, $300 to $2,500 a month |

The pattern worth noticing: nothing in that table is a platform subscription to me. The infrastructure runs in your accounts, itemized, so you own the system instead of renting it. If we part ways, everything keeps working.

## Why are AI model fees so much cheaper than people expect?

Because you pay for what the system does, not for seats or licenses. Model fees are metered per unit of work, and business tasks are small units. The clearest example I have is content:

A year of weekly SEO posts costs about $3 in model fees. The same year from an agency, at their cheapest, is $7,200. This is the economics that made AI operating systems viable for small businesses in the first place: the marginal cost of the work approaches zero, so the question stops being "can I afford to run it" and becomes "is the build worth it."

## What does the build actually cost?

The build is the real investment, and it scales with how many agents you need and how much custom wiring your existing tools require. Honest ranges from my own deals:

| What you're buying | What it covers | Investment |
|---|---|---|
| **Single-purpose automation** | One workflow: an email agent, a follow-up pipeline, a booking flow | A few thousand dollars, one time |
| **Voice agent build** | Your intake scripted, wired to calendar/CRM, tested on real calls | Same range; telephony billed to you directly |
| **Full AI operating system** | Multiple agents sharing one brain, phased in over months | Five to six figures, depending on scope |

A single automation is where most businesses should start, and I wrote a whole post on [whether that math pays](/blog/can-workflow-automation-save-money/). The phasing matters more than the sticker. A real build rolls out foundation first, then the highest-value agent, then the rest:

1. **Phase 1: Foundation** Website, data plumbing, the dashboard, the knowledge brain. The base every agent plugs into.
2. **Phase 2: Lead capture** Usually the 24/7 voice agent or missed-call automation, because it pays for itself fastest.
3. **Phase 3: Content and follow-up** The content engine and follow-up automations start compounding on top of the foundation.
4. **Ongoing: Promote to autopilot** Each agent starts gated behind your approval and earns its autonomy as you watch it work.

You see working pieces in the first month, and each phase has to justify the next. Anyone who quotes you the full system before hearing your symptoms is selling, not diagnosing.

## How do you avoid overpaying for an AI system?

Start with a diagnosis, not a purchase. My door is a **paid audit, in the low thousands, credited back 100% against the build if you proceed**. You get a map of where your hours and leads are actually leaking, a demo dashboard with your own use cases mocked up, and a phased roadmap with real numbers, whether you build with me or not. That structure exists because it keeps both sides honest: you're not committing to a big number to find out what you need, and I don't have to bury discovery costs inside an inflated build price.

And run the meter on doing nothing while you're deciding. If your business misses calls, drops follow-up, and publishes nothing, that cost is already being paid every month, it just doesn't arrive as an invoice. A plumbing company I built for was losing after-hours emergency calls, the highest-value calls it gets, right up until [the voice agent started answering them at 2am](/case-studies/instant-water-heaters/).

If you want the diagnosis version of this article for your own business, [book the 15-minute call](/book/). I'll tell you which single automation to start with, and if the honest answer is that you don't need a system at all, you'll get that answer for free.

## How Much Does an AI Voice Agent Actually Cost?
URL: https://mysickbuilds.com/blog/how-much-does-an-ai-voice-agent-cost/
Date: July 24, 2026

The owner of a plumbing company used to answer the same intake questions himself, call after call, week after week. The voice agent I built him runs his exact 13-step intake on every call now, and it [gave him back more than 20 hours a week](/case-studies/instant-water-heaters/). So what does something like that actually cost? A one-time build in the low thousands, then pennies per call to run. Here are the real numbers, line by line, next to what the alternatives cost.

- **<$1** per call in usage fees, 24/7
- **~$37K/yr** median receptionist wage (BLS)
- **168 hrs** the week your phone actually rings

## How much does an AI voice agent cost?

Two numbers, honestly separated: the build and the running costs. The build is a flat, one-time fee in the low thousands, scoped to your intake and your tools. Running it is metered usage plus modest infrastructure, all billed to accounts you own:

| Line item | What it is | What it runs |
|---|---|---|
| **The build** | Your intake scripted, wired, and tested end to end | A few thousand dollars, one time |
| **Platform minutes** | The metered cost of the agent talking | [~$0.08 to $0.16 per minute](https://elevenlabs.io/pricing/agents) |
| **Telephony + infrastructure** | Phone numbers, call routing, hosting | $150 to $1,200+ a month, usage-driven |
| **AI model fees** | The thinking behind the answers, metered | Pennies per call |
| **Maintenance plan** | Optional: monitoring, tuning, improvements | Separate, $300 to $2,500 a month |

Do the per-call math once and it stops being abstract: a typical four-minute intake call is about $0.32 to $0.64 in platform minutes plus pennies in model and telephony fees. Well under a dollar, at any hour, with no busy signal.

## How does that compare to hiring someone?

The [median receptionist wage is about $37,000 a year](https://www.bls.gov/ooh/office-and-administrative-support/receptionists.htm) before taxes and benefits, and that buys roughly 40 hours of coverage a week, one call at a time. The honest comparison isn't "fire the human": it's what each option actually delivers for the money.

The agent isn't warmer than a great receptionist, and it shouldn't make judgment calls. What it does is refuse to sleep, take two calls at once, and follow your intake perfectly on call four hundred, and it charges you cents for it.

## What are you actually paying for? Coverage.

Your business week has 168 hours in it. An office phone gets answered for about 50 of them, and [roughly 40% of appointments get booked outside business hours](/blog/can-ai-voice-agents-book-appointments/):

That grid is the real product. A dead water heater at 11pm is a same-day, high-margin job for whoever answers, and before the agent, that plumbing company's after-hours emergencies went to voicemail, which is to say, to a competitor.

## What drives the build price up or down?

Complexity and connections. A simple after-hours answering flow sits at the bottom of the range. The price climbs with the depth of the intake and the number of systems the agent writes to. That plumbing build runs a 13-step intake, quotes from a knowledge base, logs the job to the CRM, generates a work order, posts a team notification, and texts the customer and the on-call tech. Every one of those wires is build time, and every one removes a manual step the owner used to do by hand.

A good build also ships with a trust loop, and you should ask for one. On that plumbing system, every job the agent books posts a notice to the team's Slack for a quick human audit, so the owner can check what the agent said and did in seconds instead of taking it on faith. You're not buying blind autonomy; you're buying work you can inspect.

Two cost-control rules I give every client. First, the infrastructure gets billed to accounts you own, itemized, so you're never renting your own phone system back from a vendor. Second, start with the one call flow that's bleeding the most, usually after-hours or overflow, and expand once it's proven. If a voice agent is just one piece of a bigger leak, the [AI operating system cost breakdown](/blog/how-much-does-an-ai-operating-system-cost/) covers how the full-system math works.

## When is a voice agent not worth it?

When coverage is already solved. If you get a handful of calls a week and a human reliably answers them, keep your setup and spend nothing. The build pays for itself on volume and misses: calls that ring out during your busiest weeks, after-hours demand that books with whoever picks up, or an owner spending unpaid hours as the switchboard.

Here's the number to check before any of mine: how many calls did your business miss last month, and what's one booked job worth? If those two numbers multiply into more than a build costs, the phone is already deciding for you. If you want that math run on your actual call volume, [that's the 15-minute call](/book/). I'll tell you what a build would cost for your intake, and if the honest answer is "keep your receptionist and add nothing," you'll hear that too.

## Can Workflow Automation Save My Business Money?
URL: https://mysickbuilds.com/blog/can-workflow-automation-save-money/
Date: July 16, 2026

A family farm I built for used to spend two and a half hours every week on a single phone call: walking through inventory, line by line, so everyone knew what to pick and pack. One workflow automation [took that call down to 31 minutes](/case-studies/mooberry-farms/). Same information, 79% less time, every single week. So can workflow automation save your business money? Yes, and it isn't magic. It's math. This post shows you where the savings actually come from, what a build costs, and how to tell whether a workflow is quietly costing you money right now.

- **1/3 of tasks** already automatable in most jobs
- **~$13k/yr** value of saving 5 hours a week
- **1-3 months** typical payback on one build

## How does workflow automation actually save money?

Three ways. It gives you back staff hours spent on repetitive tasks, it removes the cost of fixing manual mistakes, and it captures revenue that used to slip away after hours. Most workflows hit at least two of the three, and the value stacks up week after week.

The first bucket is the biggest. [McKinsey found](https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/four-fundamentals-of-workplace-automation) that in about 60% of jobs, at least a third of the tasks can already be automated with today's technology. That third is the repeated, predictable work: copying an order from an email into your system, sending the same reminder text, chasing the same overdue invoice. You are paying real wages for it right now. The second bucket is quieter but real: [manual data entry runs a 1% to 4% error rate](https://www.lido.app/blog/data-entry-error-rates), so for every hundred fields somebody types, a few are wrong, and each one costs $10 to $100 to catch and fix. A wrong number in an invoice or an order costs time and trust to unwind. The third bucket is the one owners feel most: a lead that comes in at 9pm and gets a callback three days later is usually a lead that already hired someone else.

## What does automation replace, in real hours?

Here is what the repeated work looks like once you price it in hours. These are the tasks I automate most often for service businesses, and the weekly hours they hand back.

| Metric | Before | After | Change |
| --- | --- | --- | --- |
| Lead follow-up | 5-8 hrs/week by hand | automatic, within minutes | ~6 hrs/week back |
| Invoicing and payment chasing | 3-5 hrs/week | sent and chased automatically | ~4 hrs/week back |
| Moving data between apps | retyped by hand, 1-4% error rate | synced automatically, no retyping | fewer errors, hours back |
| After-hours leads | lost to voicemail | captured and logged 24/7 | revenue that used to leak |

None of these are exotic. They are the jobs that fill the gaps between the work you actually sell, and they are the first things to fall through when you get busy. Automating them doesn't just save the hours. It saves the hours during your busiest weeks, which is exactly when those hours are worth the most.

## What does it cost, and when does it pay for itself?

A single-purpose workflow automation is usually a few thousand dollars to build, plus a small monthly fee to host and maintain it. It pays for itself the moment the hours it saves are worth more than the build, which for most repeated tasks is a few months, not a few years.

Run the math on your own numbers. Say a workflow saves five hours a week and your time is worth $50 an hour. That's $250 a week, about $1,080 a month, roughly $13,000 a year, coming straight back to you. Against a build around $3,000, that pays for itself just before month three, and every month after is profit.

A plumbing company I built for [reclaimed more than 20 hours a week](/case-studies/instant-water-heaters/) of owner time this way. The build was a one-time cost. The 20 hours come back every week, forever.

Put your own numbers in and see what one repeated task is actually costing you.

## When does workflow automation not save you money?

When the task is rare, or never runs the same way twice. Automating something that happens once a quarter, or that needs a human judgment call every time, costs more to build and maintain than it will ever save. The honest rule: automate the work that's both repeated and predictable, and leave the rest alone.

This is where a lot of automation money gets wasted. Someone automates a flashy edge case that fires twice a year and skips the boring invoice reminder that would have saved four hours a week. If your problems compound across the whole business instead of living in one task, that's less a single-workflow question and more an [AI operating system](/blog/what-is-an-ai-operating-system/) question, and I wrote a separate honest guide on that. But for most owners, the right first move is one workflow on the one task that's bleeding the most hours.

## How do you know if a workflow is quietly costing you money?

Run through these. Each one is a place where repeated manual work is turning your time into overhead. Count your yeses.

Whatever your number was, here's the part that stings: that time is already being spent. You paid for it last week, and you'll pay for it again this week, whether or not you ever automate a thing. The only question is whether it keeps coming out of your payroll or goes back into work you can actually bill for.

If you want a second set of eyes on where your hours are leaking, that's what [my 15-minute call](/book/) is for. I'll tell you which task to automate first, and I'll tell you honestly if the answer is none, because the owners who start with one workflow and come back for the next are the best clients I have.

## Can AI Voice Agents Book Appointments Without Human Help?
URL: https://mysickbuilds.com/blog/can-ai-voice-agents-book-appointments/
Date: July 14, 2026

Yes. A properly built AI voice agent can answer a call, qualify the caller, quote the job, book the appointment, and notify your team without a human touching any of it. I know because I've built one that does exactly that for a plumbing company, around the clock. This post walks through the actual call flow, the after-hours numbers that make booking automation worth it, and the honest limits nobody selling you "AI receptionists" likes to mention.

## Can an AI voice agent actually book an appointment by itself?

It can, if "booking" is built as a complete flow instead of a chat trick. Booking means five jobs done in one call: answer, qualify, quote, schedule, and hand off. A production voice agent does all five, writes the result where your team already works, and never puts a caller on hold.

The proof I point to is [Betty, the voice agent I built for Instant Water Heaters](/case-studies/instant-water-heaters/), a water heater company in St. Louis. Betty answers every call on the first ring, day or night. She runs a 13-step intake, quotes the job from a knowledge base, logs everything to the CRM, and texts the on-call technician with the full picture when the caller hangs up. No human is in the loop unless Betty decides one needs to be.

The part most people miss: none of that intelligence was generic. We built the 13-step intake working side by side with the owner. He had a big hand in what that workflow looked like, because he'd answered thousands of these calls himself: tank versus tankless, pilot light, pressure relief valve, access to the unit. The agent asks what his best dispatcher would ask, in the order he'd ask it. That's why it books jobs instead of taking messages.

## What does the booking flow look like on a real call?

1. **Answer on the first ring** Any hour, any day. No hold music, no voicemail, and concurrent calls all get picked up.
2. **Run the intake** The same scripted questions on every call, 13 steps for the plumbing build, so nothing gets skipped at 2am.
3. **Quote and qualify** Pricing comes from a knowledge base, not a callback. Ready-to-book emergencies get separated from price-shoppers.
4. **Book and hand off** The job lands in the CRM and the on-call tech gets a text with the full intake, the moment the call ends.

What surprised the owner during testing wasn't the mechanics. It was how the calls felt. His words, reviewing the early test calls: "Love the conversation and the knowledge that she was able to pull. Overall this is very natural and much more fluid." He even liked a detail I didn't expect anyone to notice: "I can hear background noise and I love it, makes it seem more natural."

> **FROM TESTING.** The owner tried to trip the agent on purpose during a test call. His verdict afterward: "Betty handled my curveball very well. It was actually really impressive." That test-until-you-trust-it phase is part of every build, and it's exactly where a skeptical owner should spend their attention.

## Why does after-hours booking matter so much?

Because that's when the bookings actually happen. Around [40% of appointments are booked outside business hours](https://www.zippia.com/advice/appointment-scheduling-statistics/), and the same research puts the peak booking window at Sunday between 4 and 8pm. Your customers decide to book when the problem is in front of them, not when your office opens.

- **~40%** of appointments are booked after business hours
- **Sun 4-8pm** the peak booking window, per scheduling research
- **20+ hrs/wk** of owner time one plumbing company reclaimed from phone intake

For service businesses the stakes are sharper, because the after-hours call is often the *best* call. A dead water heater at 11pm is an urgent, same-day buy from a customer ready to book the moment someone picks up. Before Betty, those calls rolled to voicemail overnight, and by morning the customer had already booked the next plumber. Now they're captured and dispatched in minutes, which is exactly the [connected-system logic](/blog/what-is-an-ai-operating-system/) that separates a real automation from a gadget.

## What can't an AI booking agent do?

It can't exercise judgment it wasn't given, and you shouldn't want it to. An AI agent books the appointment types you define, follows the escalation rules you set, and should hand anything genuinely weird to a human on purpose. If a caller needs a custom commercial bid, or is upset, or asks something outside the knowledge base, the right behavior is a clean handoff with full context, not a confident guess.

Two more honest limits. First, the agent is only as good as the intake you build into it, which is why the side-by-side build with the owner matters more than the AI model does. Second, calendar complexity is real: multi-tech scheduling with hard routing constraints takes integration work during the build. None of this is a reason not to automate booking. It's the reason to build it with the escalation paths in from day one.

## How do you know if booking automation fits your business?

Simple test: count last week's calls that ended in voicemail, and ask what a booked job is worth to you. If missed calls are measured in jobs, not minutes, the math usually answers itself. What does a month of after-hours callers booking the next company actually cost you?

If you want a second set of eyes on that math, [book a 15-minute call](/book/) and bring your call volume. I'll tell you straight whether a booking agent pays for itself in your shop or whether you should start with something smaller, the same way the 13-step intake started: one owner, one whiteboard, and the questions his best dispatcher would ask.

## What Is an AI Operating System and Do I Need One?
URL: https://mysickbuilds.com/blog/what-is-an-ai-operating-system/
Date: July 9, 2026

Earlier this year I asked the owner of a roofing company doing eight figures what was holding him back from the next level. His answer: "We don't exist." An AI operating system is the fix for exactly that problem: a team of AI agents wired into one system that answers your phones, publishes your content, chases your leads, and reports back to you, around the clock. This post explains what that actually means in plain English, and gives you an honest way to decide whether you need one or whether a single automation would serve you better.

## What is an AI operating system?

An AI operating system (AIOS) is a set of AI agents connected into one system that runs real parts of your business: phone coverage, lead follow-up, content, research, and reporting. The agents share one memory, hand work to each other, and escalate to a human when a decision actually needs you.

The name is borrowed from computing on purpose. Your laptop's operating system is the layer that makes the hardware, apps, and files work together so you don't have to think about it. An AIOS plays the same role for your business operations. The phone agent that books a roof inspection writes the lead into the same system the follow-up agent reads from. The content agent that publishes a blog post pulls from the same knowledge the phone agent uses to answer questions. Nothing gets retyped, nothing gets forwarded, nothing waits for Monday.

An "agent," if the word is new to you, is a piece of AI software that does a job instead of just answering a question. ChatGPT waits for you to type. An agent watches for something to happen (a call, a form fill, a storm rolling into the county) and then does the work.

That roofing owner I mentioned is a client now, so I can tell you what "we don't exist" looked like up close. Ten million a year in revenue, built on referrals and yard signs. Zero ad spend. A 12-year-old website that one of us described as a business card with a URL. One person answering the phones, which meant a bathroom break could send a caller to voicemail, and storm-season callers don't leave voicemails. They call the next roofer on the list. The business was excellent. The system around it didn't exist.

## Why do so many businesses have AI tools but no AI system?

Because tools are easy and systems are work. The [U.S. Chamber of Commerce reports 58% of small businesses](https://www.uschamber.com/co/run/technology/ai-powered-growth-engines) now use generative AI, up from 23% in 2023. But the [JPMorganChase Institute](https://www.jpmorganchase.com/institute/all-topics/business-growth-and-entrepreneurship/understanding-ai-use-by-small-businesses) puts actual operational adoption at under 20%.

Read those two numbers together and the picture gets clear. Most owners have AI the way they have a gym membership. Somebody on the team drafts emails with ChatGPT. Somebody else summarizes a contract now and then. Useful, sure. But nobody would say AI runs any part of the operation, because it doesn't. The moment the human stops prompting, the AI stops working.

Here's the difference in practice:

| | AI tools (58% of businesses) | An AI operating system (~18%) |
|---|---|---|
| **Starts working when** | You open a tab and type | Something happens: a call, a lead, a storm, a schedule |
| **Memory** | Forgets you between chats | One shared memory across every agent |
| **Handoffs** | You copy-paste between apps | Agents pass work to each other automatically |
| **Coverage** | Your working hours | 24/7, including concurrent calls at 2am |
| **Output** | A draft you still have to use | Finished work: the post is live, the lead is in the CRM |
| **When you're on vacation** | Nothing happens | Everything happens |

I spent 25 years building and running systems inside companies like Amazon, CBRE, and Marriott before starting My Sick Builds. The thing I watched enterprises pay millions for was never a single clever tool. It was integration: systems that talk to each other so work flows without a person carrying it between departments. That's what an AIOS is. The only new part is that a roofing company can now rent that capability for less than the cost of one part-time hire, which is why I build them for contractors and service businesses instead of Fortune 500s.

## What actually runs inside an AI operating system?

There's no fixed part list; you build the agents your bottlenecks call for. But my builds usually draw from the same core modules, and seeing them in one place makes the concept concrete:

| Module | What it does | Example from a real build |
|---|---|---|
| **Voice agent** | Answers every call 24/7, qualifies the caller, books the job, logs it to your CRM | A plumbing company [reclaimed 20+ hours a week](/case-studies/instant-water-heaters/) of owner time and stopped losing after-hours emergencies |
| **Knowledge brain** | One place where your prices, policies, and process answers live, so every agent answers the way you would | Feeds the voice agent, the website chat, and the content engine from the same source |
| **Workflow automations** | The invisible plumbing: lead routing, follow-up sequences, invoice chasing, review requests | A farm [cut a 2.5-hour weekly inventory call to 31 minutes](/case-studies/mooberry-farms/) with one voice workflow |
| **Content engine** | Researches what your customers search for, writes and publishes SEO posts and city pages | A storm hits a county, and a targeted landing page is live before the door knockers arrive |
| **Research agent** | Deep-dives a question (a new market, a competitor, a vendor) and returns a report | "Should we expand into Bowling Green?" becomes a document, not a month of maybes |
| **Dashboard** | One screen showing what every agent did, with approval buttons for anything you've gated | The owner sees every lead, page, and post without asking anyone |

Two things matter more than any single module.

First, the wiring. Each module is available as a standalone product from a dozen vendors. What makes it an operating system is that they share state: the storm-monitoring agent triggers the landing-page builder, which briefs the voice agent, which writes to the CRM that the follow-up automation watches. Buy those four things separately and you own four subscriptions and a part-time integration hobby.

Here's that wiring as an actual sequence, using the storm example from a roofing build:

1. **Storm detected** The recon agent spots hail in a target county via weather data, no human watching a radar.
2. **Pages and ads go live** A landing page for that town is built and published, with geotargeted ads pointed at it, within minutes.
3. **Voice agent briefed** The 24/7 phone agent already knows about the storm, so callers hear relevant answers, not a generic script.
4. **Lead lands, follow-up starts** Every qualified call is logged to the CRM and the follow-up sequence starts without anyone remembering to.

Second, the approval tiers. "AI runs my operations" sounds terrifying until you see how control actually works. On that roofing build, we set three tiers together: things that run on full autopilot (blog posts, research reports), things that run but notify him (storm pages going live, ad adjustments), and things that wait for his explicit approval (anything customer-facing he hasn't learned to trust yet). His plan is to start most things in the approval tier and promote them to autopilot as he watches them work. That's the right instinct, and any builder who doesn't offer it is asking you to gamble.

## Do you need an AIOS, or just a few automations?

Honest answer: if one bottleneck is costing you the most, you need one automation, not an operating system. A missed-call text-back workflow or a single [custom GPT trained on your business](/blog/how-to-build-a-custom-gpt-for-your-business/) can be live in days and might be all you need this year. Anyone who prescribes the full system before hearing your symptoms is selling, not diagnosing.

The signal that you've outgrown single tools is compounding: problems that feed each other. Run through these and count your yeses.

Zero to one yes: buy the single fix for your one problem and move on. Two or three: start with the highest-cost bottleneck, but build it on a foundation that other agents can plug into later, so you're not starting over in a year. Four or more: your problem isn't any single tool. It's that nothing connects, and every gap between tools is staffed by you.

There's also a quieter reason owners build the system version, one the roofing owner raised himself. A business that runs on documented systems instead of the owner's memory is worth more. He'd been through a private-equity approach a couple of years back, and what stuck with him was how buyers price operational maturity. Neither of us is building toward a sale. But "the phones answer themselves and the marketing compounds without me" is exactly what a buyer pays a premium for, and exactly what makes a business survivable when the owner finally takes two weeks off.

## What does an AI operating system cost to run?

Less than the sticker price of the idea suggests, because running costs and build costs are different animals. The cloud servers my builds run on cost $60 to $100 a year each, and a build typically uses two: one for the website, one where the agents live. AI model fees are metered like a utility, and the meter runs slower than people expect. A researched, SEO-structured blog post costs about $0.06 in model fees on my current stack. Forty-eight posts a year, about three dollars. Agencies charge $150 to $500 for the same post.

| Metric | Before | After | Change |
| --- | --- | --- | --- |
| SEO blog post | $150 to $500 each (agency) | about $0.06 in model fees | pennies on the dollar |
| Phone coverage | one person's working hours | 24/7, handles concurrent calls | zero missed after-hours calls |
| Weekly inventory call | 2.5 hours of the owner's week | 31 minutes | 79% less time |

The build itself is the real investment, and it scales with how many agents you need and how much custom integration your existing tools require. That deserves real numbers with real context, so I'm giving costs their own post rather than a drive-by paragraph here. The short version: it's phased (foundation first, then lead capture, then content and ads), you see working pieces in month one, and nothing about it requires the budget of the enterprises I used to build this for.

## How do you figure out what your business actually needs?

Start with an inventory, not a purchase. List the work that only happens when a specific person does it: answering, quoting, chasing, posting, remembering. Price each item in hours per week, multiply by what your time actually costs, and the list ranks itself. The top one or two items tell you whether you need a $500 workflow or a phased system build. Most owners have never seen their operation laid out this way, and the exercise is worth doing even if you never spend a dollar with anyone.

Ask yourself one more question while you're at it: how many calls, leads, and search visits went to a competitor last quarter because nothing at your business was awake to catch them? Whatever that number is, it's already being spent. The only choice is whether it keeps going to the other guy.

If you'd rather have a second set of eyes on it, that inventory conversation is what [my 15-minute call](/book/) is. I'll tell you if a single automation solves it, and I'll tell you that happily, because the owners who start small and come back for the system are the best clients I have. That's how a guy with a 12-year-old website and one phone line went from "we don't exist" to a six-agent build with a storm-response pipeline. The system didn't start with a purchase. It started with a list of what was leaking.

## How to Build a Custom GPT for Your Business
URL: https://mysickbuilds.com/blog/how-to-build-a-custom-gpt-for-your-business/
Date: October 25, 2024

A custom GPT is a version of ChatGPT trained on your own business: your services, your policies, and your tone. Building one takes no code and about an afternoon, and it can answer customer questions, draft content, and handle the repetitive work that eats your week. Below is the full walkthrough, the same one from the video above.

## What is a custom GPT?

A GPT, or Generative Pre-trained Transformer, is an AI model trained on specific data. It can assist with tasks like answering customer queries, creating content, and even engaging in conversations. Essentially, GPTs are designed to interact with your business data and tasks, helping you deliver value faster. Developed by OpenAI, GPTs are a valuable superpower to have in your business.

## Why does your business need a custom GPT?

Here's why your business absolutely needs a GPT:

1. **24/7 Availability:** You'll never miss out on customer inquiries, whether they come after hours or on weekends.
2. **Cost-Efficiency:** By automating repetitive tasks, you can cut down on operational costs without sacrificing quality.
3. **Boost Productivity:** Your team will have more time for strategic, high-impact work rather than getting bogged down in daily tasks so you can move the needle and focus on growing and scaling.

In short, GPTs provide always-on support and efficiency for both your customers and employees.

## Who should your GPT be for?

Before building your GPT, it's essential to define who it's for. You'll need to decide if it'll be used internally, for employees, or externally, for customers. Knowing your audience helps you tailor the functionality and instructions for your GPT. Make sure you've got this locked in before moving forward!

## What are real-world use cases for a business GPT?

GPTs are incredibly flexible and can be customized for just about any business need. Here at My Sick Builds, I have yet to find a business that AI is unable to help. For a real example, see how I put an [AI voice agent on a farm's weekly inventory call](/case-studies/mooberry-farms/) and cut it from 2.5 hours to 31 minutes.

Here are a few real-world use cases:

1. **Internal Knowledge Base:** Quickly answer employee questions, reducing time spent searching for documents.
2. **Customer Support:** Improve response times for FAQs and troubleshooting, providing a consistent customer experience.
3. **Onboarding and Training:** Use GPTs to guide new employees through SOPs and processes.
4. **Content Creation:** Automatically draft blog posts, social media posts, or newsletters.
5. **Marketing Campaigns:** Analyze customer preferences to create more engaging content and higher conversions.

## What should you consider before building a GPT?

There are some critical factors to consider when building a GPT:

- **Data Sensitivity:** Avoid including private company or customer data. Think of it like Google: if someone has access to your GPT, they can potentially access all the data it's trained on.
- **Data Relevance:** Make sure the data you use to train the GPT is relevant to the task at hand. Keeping your information focused ensures better responses.
- **Regular Updates:** Treat your GPT like a living, breathing document. Regularly review and update your data as your business evolves to keep your GPT effective.

## How do you create a custom GPT, step by step?

Now let's get into building your custom GPT!

1. **Gather Data:** Start by collecting your data. This can be in the form of Word documents or Google Docs, which will serve as the foundation of your GPT's knowledge base.
2. **Train the Model:** Once your data is ready, you'll upload it to train the GPT. As your business grows, continue to update the model regularly (monthly or quarterly is ideal) to keep it aligned with your current operations.
3. **Refine the GPT:** Over time, you can refine the GPT's instructions and capabilities to make sure it's delivering value.

## What do you need to build a custom GPT?

Here's what you'll need to get started:

- **ChatGPT Account:** You can sign up at [www.ChatGPT.com](https://www.chatgpt.com).
- **ChatGPT Plus Plan:** This will cost between $20 and $25 per month. For most solopreneurs or small teams, the $20 plan is sufficient, but larger teams (2 or more), you'll need the $25 business plan.

## How do you build your GPT inside ChatGPT?

To build your GPT:

1. Sign into your ChatGPT account and upgrade your plan.
2. Once logged in, go to the "My GPTs" section and click on "Create a GPT."
3. Start by giving your GPT a name. I named mine "My Sick Builds Brain."
4. Add custom instructions that explain what your GPT is supposed to do.
5. Upload your documents. This could be files like brand details or meeting summaries.
6. Customize capabilities like web browsing or image generation if needed.

In just a few steps, you'll have a GPT ready to streamline your business!

## Conclusion

Creating a custom GPT can revolutionize the way your business operates. From handling customer support to generating content, GPTs save time, improve efficiency, and help you scale. Remember, the key is to tailor the GPT to your specific business needs and update it regularly as your company evolves.

Not sure where a GPT (or a bigger automation) fits your business? Start with a free [AI Opportunity Map](/ai-opportunity-map/), a no-cost scan that shows the highest-leverage places to put AI to work.

**Interested in seeing how AI can help you scale and grow your business?**

[Book Your Free AI Strategy Call Here!](/book/)

## Weekly inventory: 2.5 hours → 31 minutes, by phone.
URL: https://mysickbuilds.com/case-studies/mooberry-farms/
Date: May 28, 2026

## The receipts, before the story.

**Time saved per inventory session: 2.5h → 31 min.** From 90 minutes in the freezer + 60 minutes of someone retyping it, to one phone call. First run, no practice.

- **Annual time recovered: ~104 hrs/yr.** 2 hrs/week × 52 weeks. Ryan now gets **two and a half work weeks** back.
- **People in workflow: 3 → 1.** Helper and mother-in-law both freed up.
- **Steps in workflow: 4 → 1.** One phone call. Same numbers, touched once.

> "She sounds like my banker, which I can't really get over."
>
> *Ryan · Owner, MooBerry Farms*

## Who, what, and how long.

- **Client:** MooBerry Farms
- **Industry:** Direct-to-consumer cattle farm
- **The ask:** Eliminate the 2.5-hour weekly inventory bottleneck.
- **What we built:** A custom voice agent ("Betty") that takes inventory by phone and syncs to Google Sheets + Square POS.
- **Stack:** ElevenLabs · n8n · Google Sheets · Square POS · Email
- **Timeline:** A few weekends

## A clipboard, a mother-in-law, and 2.5 hours every week.

Ryan runs a family cattle farm in Tennessee. He breaks down whole cows into thirteen different cuts and sells direct through a farm store and an online Square shop. The product is excellent. The back office, until recently, was a brick wall.

Every week, inventory went like this. Ryan stood in the freezer and weighed every cut, one by one, reading each weight out loud. A second person wrote every number down on *whatever paper was nearby*: sometimes a clipboard, sometimes the back of a napkin, sometimes the back of an envelope.

Then that paper went to his mother-in-law. She typed every weight into an Excel spreadsheet. That spreadsheet fed Square POS, which generated the purchase order. **Three people. Four steps. The same numbers, touched twice.**

> "We always gravitate back towards paper and pencil. That's how we rock. We count, we write down, we go to a computer, we type it again into a spreadsheet."
>
> *Ryan · MooBerry Farms*

### Pain points

- Same numbers entered twice: once on paper, once in Excel.
- Three people locked into the workflow every single week.
- Yield % unknown for a day or more after inventory.
- A single typo could silently corrupt Square inventory.
- Workflow broke entirely whenever someone got sick.

**Cost of inaction: ~104 hrs/yr.** 2.5 hrs every week. Three people. Half a workday gone, every single week, before anything ever made it onto the storefront.

## Meet Betty, the voice agent who answers MooBerry's phone.

Betty lives behind a single phone number. She answers in a warm, slightly Southern voice. She is trained specifically for the way Ryan does inventory: meat types, box numbers, weights. And she keeps up with him as fast as he can talk.

When Ryan finishes a box, Betty repeats the count, the total weight, and asks if it is correct. When he finishes the session, she gives him a grand total and tells him she's pushing everything to Google Sheets. There is no app. There is no spreadsheet to open. There is no second person typing.

- **Answers the phone** in Ryan's voice: warm, Southern, conversational. Tuned for noisy freezer audio.
- **Captures structured data** in real time: cut, box, weight, from natural speech. No app, no form.
- **Confirms every box** out loud before moving on. Self-correcting on the call.
- **Writes to the systems Ryan already uses:** Google Sheets, Square POS, email summary. Automatically.

## Hear Betty answer the phone.

A short clip from Ryan's first inventory call. Listen to how Betty handles the dictation, confirms the box, and catches the 0.80-vs-80 mistake in real time.

## Four steps Ryan sees. Everything else, in the background.

1. **Ryan calls Betty.** One phone number. Speaker on, freezer door open. Betty picks up: "MooBerry Farms, this is Betty. Are you ready to record some inventory?"
2. **Ryan dictates a box.** Meat type, box number, then every weight off the scale. As fast as he can talk. Betty keeps up. No app, no form, no fields.
3. **Betty confirms.** After each box: "That's 6 ribeyes, 4.81 pounds total. Sound right?" She catches mishears in real time. Ryan corrects on the call.
4. **Betty closes out.** When inventory is done: "Total weight recorded today is 183.58 pounds. I'll send this to the Google Sheet now." That is the whole workflow.

### Four things Betty is doing while Ryan talks.

- Every weight written to Ryan's existing Google Sheet: same shape, same columns, untouched.
- Summary email to Ryan with totals by box, by cut, and a grand total for the session.
- New inventory pushed to Square POS so the online storefront matches the freezer.
- Full transcript saved for every call: error-traceable, audit-friendly.

## Ryan said "0.80 pounds." Betty heard "80 pounds."

Halfway through the first call, Ryan asked Betty to add one more chuck eye steak to the previous box. He said the weight was "0.80 pounds." She heard "80 pounds."

On the old paper system, nobody would have caught it. The error would have rolled through the spreadsheet, into Square, and onto the storefront. Customers would have seen impossible inventory levels for days.

Instead, Ryan looked at the running total Betty was building, saw the number was wrong, and corrected her in real time. She updated the entry on the spot. The rest of the inventory kept moving forward, untouched.

> "We corrected it in real time, and it didn't throw off the rest of the spreadsheet. **That was incredible.**"
>
> *Ryan · MooBerry Farms*

## The full flow, start to finish.

One phone call in. Three systems updated. The dashboard MSB operates on Ryan's behalf: every call's status, impact, and history.

**Impact since May 2026:**

- **People in workflow:** 3 → 1
- **Inventory time:** ~31 min, first run
- **Data lag:** real-time, vs 3 days

## Every metric, side by side.

- **Time per inventory session:** ~2.5 hours → ~31 minutes, first try.
- **People required:** 3 → 1.
- **Steps in workflow:** 4 → 1. One call.
- **Time until yield % is known:** a day or more → within 15 minutes, same evening.
- **Risk of typo corrupting Square:** high and silent → self-correcting, caught on the call.
- **Workflow if someone is sick:** stops entirely → unaffected, 24/7.

## Inexpensive to run. Built to last.

No enterprise software. No SaaS sprawl. Just one phone number that does the work of several people, running on tools that stay cheap to operate at the scale of a small farm.

- **Voice layer · ElevenLabs.** Custom persona ("Betty"), tuned for noisy freezer rooms and cooler-fan audio.
- **Extraction · OpenAI: ChatGPT.** Turns each transcript into strict JSON. Normalizes spoken numbers and reconciles every box, cut, and weight into clean totals.
- **Orchestration · n8n.** Parses every transcript. Extracts cuts, boxes, and weights. Routes structured data.
- **System of record · Google Sheets.** Ryan's existing sheet: same shape, same columns. Nothing on his side had to change.
- **Notifications · Email.** Summary with totals by box, by cut, and a grand total. Delivered within minutes of call end.
- **Inventory sync · Square API.** Pushes new counts to the storefront: online shop always matches the freezer.
- **Reliability · Live error correction.** Self-correcting during the call. Full transcript backup of every conversation.

## Ryan's words.

Recommend to another farm? **Without question.**

> "Cory, with My Sick Builds, came to us. He's a customer of ours. He came over to the store one day and said, where are your biggest pain points? What do you do repetitively that you don't like doing?"
>
> "Every week, it's the inventorying of meats. He built Betty to receive our inventory and eliminate a pain point. The worst part of our job, other than snowstorms. **He fixed that problem.**"
>
> "She always shows up for work. She doesn't get sick. It can be 10 o'clock at night and she's going to answer that phone."
>
> *Ryan · Owner, MooBerry Farms*

**For anyone still on the fence:**

> "Keep your skepticism, but don't let it stop you from buying back time. Use technology to save you time so you can do things you enjoy more."

## Every call answered, qualified, and dispatched. 24/7.
URL: https://mysickbuilds.com/case-studies/instant-water-heaters/
Date: August 15, 2025

## The receipts, before the story.

Instant Water Heaters lives or dies by the phone, and the phone was losing.

- **24/7/365** phone coverage
- **13 steps** on every call
- **20+ hrs/wk** owner time reclaimed
- **0** calls to voicemail

> "Over 20 hours a week, just answering basic intake questions."
>
> *Owner · Instant Water Heaters*

## Who, what, and how long.

- **Client:** Instant Water Heaters
- **Industry:** Plumbing and water heater install and repair, St. Louis
- **The ask:** Stop sending calls to voicemail, and stop losing the after-hours emergency jobs.
- **What was built:** A custom 24/7 voice agent (Betty) that answers the phone, qualifies every caller with a 13-step plumbing intake, quotes the job, logs it to the CRM, and texts the on-call technician.
- **Stack:** Custom Voice AI, OpenAI, Make.com, Airtable, Twilio, Linq, Google Docs, Slack
- **Timeline:** A few weeks, then tuned on real customer calls

## The phone was the bottleneck.

Instant Water Heaters installs and repairs water heaters across St. Louis. The work is high-margin, and the emergencies are the best of it. A dead water heater is an urgent, same-day buy. The customer is ready to book the moment someone picks up.

The problem was that nobody could always pick up. At peak, calls stacked up and rolled to voicemail. After hours, the emergencies that should have been the easiest yes went unanswered until morning, by which point the customer had already called the next plumber. And the overflow that did get answered landed on the owner, who was personally absorbing hours of the same basic questions every week.

> **The cost of a missed call.** Every call that rolled to voicemail was a same-day, high-margin job, handed straight to the next plumber in the search results.

### Pain points

- Calls going to voicemail during peak hours. Every missed call was a job handed to a competitor.
- After-hours emergencies, the highest-margin work, going unanswered overnight.
- Inconsistent intake. Details got missed, and technicians arrived unprepared.
- The owner spending 20+ hours a week answering the same basic questions by hand.

## Meet Betty, the voice agent who answers the phone.

**Betty** (Instant Water Heaters · Intake Agent)

Betty answers on the first ring, day or night, in a warm, natural voice. She is trained on the way Instant Water Heaters actually does intake: tank versus tankless, pilot lights, pressure relief valves, the questions a sharp dispatcher would ask.

She runs a 13-step script on every single call, so nothing gets skipped no matter how busy it is.

What the agent handles:

- **Answers 24/7/365.** Nights, weekends, holidays. No voicemail, ever.
- **Speaks plumbing.** Tank versus tankless, pilot light, pressure relief valve, the words customers actually use.
- **Runs a 13-step intake.** Name, address, heater type, the problem, and access details. Every call.
- **Quotes on the spot.** Accurate pricing from a knowledge base, not a callback.
- **Qualifies the lead.** The ready-to-book emergency from the casual price-shopper.
- **Handles the money talk.** Takes the awkward payment conversation off the owner's plate.

Sample call:

> **Betty:** Thanks for calling Instant Water Heaters, this is Betty. What's going on with your water heater today?
>
> **Caller:** It's leaking from the bottom, pretty bad.
>
> **Betty:** Got it. Is it a tank or tankless unit, and about how old?

## One call in. A work order out.

From the caller's side it is one phone call. Behind it, Betty qualifies, logs, and dispatches, then keeps the job moving until a tech is on the way.

![Inbound phone call to Betty the AI voice agent, then intelligent qualification, instant data entry, work order creation, and job dispatched](/images/case-studies/instant-water-heaters/high-level-diagram.webp)

1. **The call comes in, 24/7.** Betty answers on the first ring. No hold music, no voicemail.
2. **Betty qualifies.** She runs the 13-step intake, captures the heater type and the problem, and quotes the job from the knowledge base.
3. **Everything lands in the CRM.** Name, address, problem, and access notes write straight into Airtable. No retyping, no missed fields.
4. **A work order is created.** Betty generates the job and a shareable work-order doc the team can open.
5. **The on-call tech gets a text.** An SMS goes out with the full job detail, so the plumber shows up prepared.
6. **The job is dispatched within minutes.** Start to finish, while the old system would still have been ringing.

### What Betty handles while the caller just talks.

- Logs the full conversation history for every call, so nothing is lost and every job is auditable.
- Texts the customer to send photos of the unit, then processes and attaches them to the work order.
- Posts a notice to the team in Slack for a quick human audit before dispatch.
- Follows up automatically if a job sits without an update, so nothing falls through the cracks.

## The whole system, not just the call.

Betty is the front door, but the build runs the entire job behind her. Voice calls, the website chat widget, and inbound texts all flow into one routing layer that logs every conversation, checks authorization before anything moves, spins up the work order, loops in the team, and handles the customer's photos, all automatically.

![Full technical workflow for the Instant Water Heaters voice agent: voice, website chatbot, and iMessage or SMS all route through a logging and authorization layer that creates a Google Doc work order, posts a Slack notice for a team audit, and texts the customer for photos and the plumber for dispatch](/images/case-studies/instant-water-heaters/technical-workflow.webp)

## What changed.

The receipts up top are the headline. Here is the rest of what moved, the parts a caller never sees.

| Metric | Before | After | Change |
| --- | --- | --- | --- |
| Phone coverage | Business hours only | 24/7/365 | always on |
| Calls at peak | Rolled to voicemail | Every line answered | 0 missed |
| After-hours emergencies | Lost overnight | Captured and dispatched | in minutes |
| Lead qualification | Whoever answered guessed | Same 13 steps, every caller | consistent |
| The money talk | An awkward callback | Handled on the first call | up front |
| Job detail to the tech | A name and a number | Full intake, plus photos | prepared |
| Stalled jobs | Quietly forgotten | Auto-followed-up | within the hour |
| Owner time on intake | 20+ hours a week | Handed off to Betty | reclaimed |

[Audio: Hear Betty adapt on a second call](/audio/case-studies/instant-water-heaters-call-02.mp3)

## Inexpensive to run. Built to last.

No enterprise call center. No new phone system to learn. Just one number that answers every time, running on tools that stay cheap at the scale of a local plumbing company.

- **Voice layer: Custom Voice AI.** A custom real-time voice agent (Betty), tuned for plumbing intake and natural phone conversation.
- **Language: OpenAI.** Turns what callers actually say into clean, structured intake.
- **Orchestration: Make.com.** Routes every call, logs the history, and runs the follow-up and dispatch logic.
- **CRM: Airtable.** System of record for every job, with clean fields and zero double entry.
- **Telephony and SMS: Twilio.** The phone line, the technician alerts, and the photo intake.
- **Customer texting: Linq.** Carries the customer contact and the photo-request text thread.
- **Work orders: Google Docs.** A shareable work-order doc per job, ready for the field.
- **Team alerts: Slack.** A quick human audit step before a job goes out the door.

## The turn.

For a home-services business, the phone is the front door. When it goes to voicemail, the job walks next door. Betty makes sure it never does. She answers every call, asks every question, quotes the work, and gets a tech moving, at 2pm or 2am, without the owner ever picking up.

If the phone is the bottleneck in your business, that is exactly the kind of thing I look for first.

## 144 hours a year of guest messages, handled.
URL: https://mysickbuilds.com/case-studies/airbnb-chatbot/
Date: May 1, 2024

## The receipts, before the story.

A custom chatbot quietly handed Josh his evenings back.

- **144 hrs** recovered a year
- **$14,400** saved a year
- **24/7** guest coverage

> "You need to get on board with this. You don't want to be pulling your hair out, answering those repetitive questions day after day, night after night."
>
> *Josh · Airbnb host*

[Video: Josh's review of the Airbnb guest chatbot](https://www.youtube.com/watch?v=D_c7MIXfJ1I)

## Who, what, and how long.

- **Client:** Josh, an Airbnb host
- **Industry:** Short-term rental and hospitality
- **The ask:** Stop the around-the-clock guest messaging from eating into his life.
- **What was built:** A custom 24/7 chatbot that answers guests in Josh's own voice, handles the repetitive questions automatically, and escalates only the ones that need him.
- **Stack:** Custom chatbot, FAQ knowledge base built with Josh, tiered human escalation
- **Timeline:** A few weeks, then refined on real guest conversations

## The phone never stopped buzzing.

Josh hosts on Airbnb. Great property, great reviews. The part that did not scale was the messaging. Guests asked the same handful of things over and over, check-in instructions, the wifi password, where to park, what to do nearby, and they asked at every hour of the day and night.

> **The real cost.** A 2am "how do I get in?" is a small emergency for the guest and a real interruption for the host. Josh was always on call, and the questions were almost never new.

### Pain points

- The same questions, over and over: check-in, amenities, recommendations.
- Messages arriving at every hour, including late night and early morning.
- An always-on expectation that never actually switched off.
- Real stress and lost personal time, week after week.

## The build: a chatbot that sounds like Josh.

Three things made this work instead of feeling like a generic bot.

- **It sounds like Josh.** I programmed it with his communication style, so guests get his warm, friendly tone, not a robotic FAQ.
- **It actually knows the answers.** I built a comprehensive FAQ database with Josh, covering the questions guests really ask, from check-in to the wifi to the best coffee down the street.
- **It knows when to step aside.** A tiered response system handles the routine and escalates anything complex straight to Josh. Guests never hit a dead end.

It runs 24/7 and answers the moment a guest message comes in.

## How a guest conversation flows.

1. **A guest messages.** Any hour, any question.
2. **The chatbot answers in Josh's voice.** Check-in steps, wifi, parking, local recommendations, all instantly.
3. **Routine questions never reach Josh.** The repetitive majority is handled without him ever touching his phone.
4. **Anything complex escalates.** The chatbot hands the rare, tricky message to Josh with context, so he steps in only when it counts.

## The numbers, twelve months in.

The chatbot compounded quietly. Every month it absorbed about twelve hours of messaging Josh used to do himself. Valued at his own time, that is around $1,200 a month, and over a year it adds up to real money and a real chunk of his life back.

| Period | Time saved | Cost saved |
| --- | --- | --- |
| Per month | ~12 hours | $1,200 |
| Per 6 months | ~72 hours | $7,200 |
| Per year | 144 hours | $14,400 |

## Always on, cheap to run.

For a host, the property is the product, but the messaging is the job. This one quietly does the job, at every hour, so Josh does not have to.

- **Chatbot: Custom Chatbot.** Answers guests in Josh's own voice, in seconds, 24/7, including the 2am check-in question.
- **Knowledge: FAQ Database.** Built with Josh. His real answers to the questions guests actually ask, not a generic script.
- **Escalation: Tiered Handoff.** Routes anything complex straight to Josh, with context, so a human steps in only when it matters.