Where Should the Agent Act? New Approach to Product Discovery for Agentic Systems
summary
Claims that product discovery must find where time actually goes and place agents at specific judgment points, not chase model demos. Method: shadow real work, map hesitation and handoffs, break tasks into agentic steps, prototype agents that extract, validate, draft and escalate. Example: environmental product declarations move from months to hours with higher consistency. Small cross‑functional teams iterate with evals as real cases and reinvest saved capacity into new product value.
AI-written summary · about
chapters (12)
- 0:00Agentic systems introduction
- 3:04Human agent in workflows
- 7:20Scaling expert judgment
- 11:54Efficiency vs capacity
- 14:25Agentic workflow examples
- 17:51Product discovery for agents
- 23:11Learning from expert feedback
- 25:21Team roles in agentic products
- 28:09Leadership in adoption
- 30:36New opportunities with agents
- 34:10Starting agentic discovery
- 35:14Closing and call to action
related talks
33mAgent Experience
21mThe Agent Native Company
30mWhy Agents Don’t Ship: Closing the Production Gap for Real Business Impact
28mAI Agents as Architectural Strategists | Aligning System Design with Business Goals
15mBest Practices and Lessons Learned in Sourcing
21mDesigning in the age of the 10x engineer
42mAgents of Change: From Time Saved to Intelligence Activated
30mAgents Building Agents