About

Who you’d be working with

I’m Christian Campbell. I build AI systems that do work people are doing by hand, and then I measure them against the people who were doing it.

That’s the whole job. Somebody’s team runs the same script, fills the same form, answers the same question, forty times a day. I take one of those, build something that does it, run it on the real work for two weeks, and put its numbers next to the team’s. Sometimes the answer is that the machine shouldn’t do it. That’s a real answer and you still keep the build.

If that’s what you’re here for, the front page explains how it works and there’s a form on it.

Why I’m the one to do it

Thirteen years as a backend engineer, the last few of them on AI systems that went to production and got measured.

• I built a conversational agent that qualified and booked recruits by text for a national direct-sales recruiting operation. It ran the same script the call center ran, across about 2,500 recruits, and booked around 74% — roughly 7 points better than the human team.
• I took a list of 523 leads a business had written off and got 17 qualified buyers back out of it, for about $200 of model cost. The agent found and qualified them. Her team closed them.
• Before that: an analysis pipeline that read thousands of support conversations and surfaced the six UX problems actually driving them.

The detail on all of it, plus client sites and side projects, is on Other Work.

How I work

Ship things that work, then iterate. No over-engineering, no building for requirements nobody has yet.

1. Understand the actual problem, not the assumed one.
2. Sit with the person who does the job and learn the real version of the process — the judgement calls, the exceptions, the part nobody wrote down.
3. Build the simplest thing that does it.
4. Measure it against the humans, honestly, including when it loses.

The best code is code you don’t have to debug at 2am, so I optimize for clarity and for getting alignment before anybody writes anything.

Background

• ConveYour — AI systems in production: conversational agents, RAG, analysis pipelines.
• HighLevel — Senior Backend Engineer on a SaaS platform with 80+ clients and millions of users.
• Tyler Technologies — criminal justice software serving government agencies.
• Years of freelance and contract work building custom business systems before that.

Stack: PHP/Laravel, Python, PostgreSQL, AWS. Claude and OpenAI APIs, RAG pipelines, BERTopic, sentence-transformers. Cloudflare Workers, Pipedream, Docker.

What I’m building now

ManagerTools — AI phone answering for small businesses. It takes the call, answers from a knowledge base, and books the appointment.

The free Proof Runs feed this. I’m doing them partly to find the next product worth building, which is why they’re free and why I take one at a time.

Let's talk

Tell me what your team does by hand. A couple of sentences is plenty — what the job is, how often it happens, and roughly what it costs you. I’ll tell you straight whether it’s a fit.