I'm building the bridge between solid data engineering and AI agents that can actually reason over your business data — and looking for 1–2 design partners to build the first real version with.
Apply as a design partnerImagine typing this into Slack:
@agent what's driving the recent growth in Q3? Give me the main reasons.
And getting back a real answer — grounded in your actual data warehouse, not a guess — because the agent can explore your dimensions, pull the right numbers, and combine them with business context it actually understands.
That's Company Brain: a governed semantic layer and data warehouse, connected to conversational AI agents, so anyone on the team can ask business questions in plain language and get answers they can trust.
Two things had to mature separately before this became possible:
BigQuery, dbt, domain-driven data marts, a governed semantic layer — the classic analytics engineering stack, done properly.
Model Context Protocol, collaborative agent workspaces (I've been building with Buzz), and text-to-SQL systems that are finally good enough to be trusted with real business questions, not just demos.
Most companies have one side or the other. Almost nobody has both, connected. That's the gap.
I'll say this plainly: this isn't a finished product, and I won't pretend it is.
I've built a working prototype — dbt, GCP and Buzz — and I know the architecture holds up. What I don't have yet is a second real-world environment to prove it in, with real data, real edge cases, and a real team asking it real questions.
That's what a design partnership is:
If you need a guaranteed outcome and a fixed spec, this isn't the right engagement — Fractional Head of Data is. This one is for teams okay with "let's find out together."
This is an application, not a mailing list signup — tell me enough that the first call is useful.
Prefer email? jice@lavocat.name · LinkedIn
Looking for something more foundational first? → Fractional Head of Data