You paid for software nobody uses.
The logins sit unused. Your team went back to typing orders in by hand, the same way as last year.
Don't get left behind. I build AI employees into your company, replacing manual tasks, and teach your team to run them.
For companies with $5M to $150M in sales
Evan Knox
Grew and sold 5 businesses
The logins sit unused. Your team went back to typing orders in by hand, the same way as last year.
More leads won't help if quotes take 3 days to go out. Work piles up at the slowest step, and that step sets your pace.
Your best rep found a faster way to use AI. It lives in their head, so the rest of the team never gets it.
Right now I'm building AI inside a $125 million Inc. 5000 company. It sorts the daily orders, writes a proposal for each customer and tells each sales rep who to call and why.
Evan Knox, AI for your business
Each AI employee does one job your team does by hand today. Each month, the jobs we hand off often add up to the work of one or two roles. You decide whether that frees people up or means fewer hires. I've built every job below.
Sorts the day's orders and flags the ones a person must check.
Builds a custom proposal for each customer, ready for a rep to review.
Gives each sales rep a daily list of who to call and why.
Answers your phone and books jobs on your calendar.
Answers customer emails using your company's own past answers.
Logs every call and email in your customer list, so nobody types it in.
Example timeline. We start wherever your slowest step is.
We fix one slow step at a time. Each month you see your numbers, before and after.
On our first call I ask for your numbers at each step, from the first time someone finds you to the day you get paid. The worst number is where we start. No numbers yet? We measure first.
Say order checks are slow. I ask how to make them faster, then ask again, until each answer is a job someone on your team does by hand today.
An AI employee is software that does one job, like checking every order against your price list. Some jobs stay with a person, and anything that doesn't match gets flagged.
It runs in a Claude account your company owns (Claude is an AI tool, like ChatGPT), connected to your email, customer list and order system. It runs on its own, on a schedule or each time new work comes in.
Each AI employee gets tested on your real work before anyone relies on it.
A training day to start, then a weekly team call. Your best person's way becomes everyone's.
AI can give a great answer once and a bad one the next time. So I run the same real orders through each AI employee many times and score every answer. Your team gets it once the scores hold up. When I change one, or a new AI model comes out, the tests run again.
Most teams have one person who's good with AI. Teaching your team to use AI spreads what that person knows to everyone, every week.
Each person brings a job from their own week and leaves with it running faster.
Whoever got the best results that week shows everyone how they did it.
I save their way in your Claude account as step-by-step instructions anyone on the team can run.
You don't rent these from me. They live in a Claude account your company owns, connected to the tools you already use. If we stop working together, they keep running in your account.
Works with the tools you already use
If your team does it on a computer or a phone, I can build it. I only build what speeds up your slowest step.
Your investment each month
$10K
Stop any month.
Estimated salary savings each year
$75K
for each role an AI employee replaces
Potential return
7x
when one month's $10K replaces a $75K yearly salary
No long contracts. No notice period. Email me any time, and it ends with the month you paid for.
It's built in your accounts and runs with or without me. There's nothing to hand back.
No account manager and no handoff. The person on your first call is the person building your AI employees.
Everything runs in a Claude account your company owns. Anthropic doesn't use business accounts' data to train its AI unless you opt in. Before we connect anything, we decide what each AI employee can read. You can remove my access any day.
Me, finding your slowest step, building and testing your AI employees, your team's training day and the weekly team call. Claude logins are paid separately, straight to Anthropic, the company behind Claude.
A Claude login for each person who uses it, about $25 a month each, paid straight to Anthropic.
Owners of US companies with $5 million to $150 million in yearly sales. Distributors, manufacturers, IT and engineering firms, contractors and freight brokers. Also home services companies with large jobs, like roofing or HVAC.
No. I build your AI employees on Claude, and your team can keep ChatGPT for anything else.
I test each AI employee before your team relies on it, and I keep checking the results while we work together. Anything that doesn't match, like an order with the wrong price, gets flagged for a person on your team.
Then it isn't working yet. Training uses each person's real work, so they leave with a job that's already faster.
Often within the first month. The first call names your slowest step and the first jobs, and those get built and tested before anything else. If that step isn't measured yet, measuring comes first.
We check what can connect on the first call. If the slow step sits in a system we can't reach, I'll tell you before we pick the first job.
Everything I built stays in your accounts and keeps running, and your team keeps every shortcut. Your company keeps paying for its own Claude logins. My checking stops when we stop, so your team takes that over.
30-minute call with Evan Knox
We walk through how work moves through your company, and you leave with a plan like this one.
If AI isn't the answer, I'll tell you.