Building the Next Generation of Software
summary
Company shifts everything: strategy, pricing, name change and launches Finn to replace Intercom; revenue climbs to $400M with Finn $100 million business. Team builds agentic features using Claude and Claude code, pays tens of millions in token spend, and reports tripled engineering productivity while 94% of production handled by Claude code. Designers must ship to production by set deadlines, 5–10% of staff opt out, and reliability math (multiplying 95% step reliabilities) forces many features back to human‑in‑the‑loop.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 3:20Going all-in on AI transformation
- 7:34How AI changes product design
- 13:52Building software in the AI era
- 20:06Team momentum and productivity gains
- 25:42No product moats in the AI age
- 26:44Q&A on driving change and measuring productivity
related talks
39mHow Intercom Doubles Down on AI
18mDispatch from the Future: building an AI-native Company
27mPixels and prompts: building products in the AI era
18mBuild like it's 2030. Listen like you mean it.
14mAgents for Everything Else
22mHow Building with AI Can Double the Throughput of Your Engineering Team
53mFireside chat on Future Developer Experience
25mFrom Analytics to Agentic Apps