Who this is for:
CTOs, founders, VPs, and heads of engineering. Teams of five or five hundred. You have a decision in front of you that's expensive to get wrong, and the people closest to it are too close to it.
What I'm most useful on:
· GenAI and agentic adoption. Whether your pilot will survive contact with production, and what the 5% that ship do differently. Model choice is almost never the problem. Workflow integration usually is.
· Platform and CI readiness. Agent-generated code commits at a volume your pipeline was not designed for. Most teams discover this after it breaks.
· Engineering org design. Squad structure, delivery model, hiring bar, and why shipping has slowed even though headcount went up.
· Build versus buy, and architecture direction. Vector stores, memory design, orchestration, guardrails and governance.
How it runs:
You send context beforehand and I come prepared. The session is working time, not a pitch. I ask questions, challenge the framing where it needs challenging, and give you my actual read rather than a menu of options. You get written recommendations within 48 hours: what I'd do, in what order, and what I'd stop doing.
Most engagements continue from here. If the work needs a longer arrangement, we scope it properly outside this platform, on a contract and an invoice.
Why me:
I lead engineering at a global research and advisory firm: strategy, platforms, hiring, and our GenAI Center of Excellence. I built the founding engineering team there and shipped the first AI-integrated products. I write about agentic systems and platform engineering at prateek-sharma.com, so you can assess how I think before spending anything.
Happy to work under an NDA. Ask before you send anything sensitive.