AI CONSULTING · SAN DIEGO COUNTY
AI consulting for
San Diego software teams.
Your engineers are being told to use AI coding agents, and nobody can say what good looks like. I help software teams work out where agents earn their place, where they don’t, and what has to be true before you scale them.
Based in Oceanside, working in person across San Diego County, and remotely everywhere else. The work happens in your repositories and your infrastructure, not in a slide deck.
What an engagement covers
Readiness assessment
A written review of how your team builds today: repositories, review culture, CI, and where agent work is already happening. You keep it whether or not we go further.
A pilot you can measure
One team, one set of repositories, a defined outcome. Success is measured in merged pull requests, not in a demo.
Rollout and governance
Wider adoption with the rules that keep it safe: which agents may change what, how their work gets reviewed, and how anyone can tell it was good.
A good fit if
—
Your team ships software and writes its own code
—
AI coding tools are already in use, or being pushed from above
—
You want a plan you can check, not a slide deck
—
You’d rather fix how you deliver than buy another tool
Probably not if
—
You want a chatbot or marketing automation
—
You need AI strategy for a business that doesn’t build software
—
You’re looking for a staffing agency
—
You want a model trained from scratch
Brett Lee
Building software since 2001 — Apple, PlayStation, Pearson, and the startups in between. I also build CTOsphere, tooling for software lifecycles where agents write much of the code, so the advice comes from running agents in real delivery work.
Questions people ask
What does an AI readiness assessment produce?
A written report covering where AI can save your team time or money, where it won't, what has to be true before you scale it, and the controls needed for security and quality. It includes a recommended first pilot with a measurable outcome.
How long does the assessment take?
Two to three weeks from kickoff, with two to four working sessions with your team.
Do you work with businesses outside software?
Yes. Most of my clients are software teams, but the same assessment works for any business with repetitive processes and the data to improve them. The report is written in plain language for owners, not engineers.
Will you recommend specific AI tools?
I'm vendor-neutral. I recommend based on your existing systems, data, and risk profile, and I say plainly when a tool isn't worth adopting.
How do you measure a pilot?
Against the outcome agreed at the start: merged pull requests, hours saved per week, cycle time, or error rates. If the pilot doesn't move the number, I say so and we stop.
Building something that needs to hold up?
Start with a conversation. No deck, no discovery phase — just the actual problem.