Practice · The Offer
/ 01 — Currently shipping · Booking Q3 pilots
AI Lab
I build AI coaching agents for Product Owners, PMs, and Scrum Masters — open-weight LLMs (Llama), CrewAI orchestration, native hooks into Jira, Azure DevOps, and ProductBoard. Trained on 17+ years of live cohort transcripts, including the May 2026 CSPO classroom.
What I ship
- →Working agents wired into your stack — not slideware.
- →Eval sets and red-team scripts so quality is measurable.
- →PO / PM / SM coaching so the team owns the agent after I leave.
Off the table
- ✕Generic chatbots with no workflow hook.
- ✕Closed-model lock-in without an exit plan.
- ✕Agents that bypass governance in regulated estates.
How an engagement runs
- 012 weeks
Discover
Workflow audit across one squad. Map the 3–5 ceremonies, artefacts, and decisions where an agent earns its keep. Output: a scoped pilot brief and a no-go list.
- 024 weeks
Pilot
Ship one agent into the live workflow — Jira / ADO / ProductBoard hooks, evaluation set, human-in-the-loop checkpoints. Output: working agent + weekly metric review.
- 038+ weeks
Embed
Roll the agent to 3–10 squads. Coach POs, PMs, and SMs on prompt patterns, eval discipline, and ownership transfer. Output: in-house capability — not vendor lock-in.
§ Contact
Let's build.
Tell me about your team, the workflow, and the one thing you'd want an AI agent to handle differently.