Talks I've given.
I speak occasionally about audio, recommendations, and product craft. Email me if you'd like one for your event.
Panel: Build vs. Buy vs. Embed — How CPOs Are Making AI Investment Decisions
Every CPO is facing the same pressure: move fast on AI without backing the wrong horse. This panel brings together product leaders who've been through the decision-making process — building proprietary models, buying third-party tooling, or embedding AI into existing workflows — and interrogates what they got right, what they'd do differently, and how they're thinking about the next wave of decisions.
Claude Code for Product Managers
AI-Powered Product Management: Best Practices for Requirements Gathering
Past talks.
From Prompts to Systems: What 6 Months of Running OpenClaw Agents Taught Me
Insights from six months running OpenClaw, an open-source personal AI system used continuously rather than as a prototype: the orchestration layer matters more than the model, memory and evolving instructions compound capability and alignment, and sustained real-world use reveals reliability issues that demos conceal. On the shift from AI as tools you use to systems that work for you.
From Prompts to Systems: What 2 Months of Running OpenClaw Agents Taught Me
Three learnings from two months running OpenClaw agents in production: the orchestration layer matters more than the model, memory and instructions compound system value, and long-term operation reveals what demos conceal. Includes live demos, where it still breaks, and what the agentic shift means for enterprises.
AWAAI Learn & Lead with Ashish Bhatia
A conversation on career growth, leadership, and building AI products — from citizen-developer tools at Microsoft to conversational discovery at Audible.
Understanding Agent Failure Modes: Lessons from Real-World AI Systems
A look at how AI agents actually break in production — and what that reveals about building systems people can trust, drawn from real-world deployment rather than demos.