Scaling AI Adoption Across 3000+ Developers at Booking
summary
Booking.com introduces SourceCraft Kodi and multiple GenAI tools across 3,000+ developers, grows from <10% to 75%+ engineers using GenAI with 65% daily users, and correlates heavy use with 30% more PR throughput plus ~16% higher innovation rate and lower change-fail percentage. Centralized procurement, targeted workshops (two-day training + 3–5 day EBAs), measurement with DX metrics, and practical guardrails on sensitive code produce measurable speedups in modernization and improved code maintainability.
AI-written summary · about
chapters (11)
- 0:00AI adoption at Booking.com
- 1:28Booking scale and challenges
- 4:22Three pillars for AI enablement
- 6:07Early experiments and measurement
- 8:55Developer adoption barriers
- 11:42Building trust and concerns
- 13:27Training and community building
- 15:13Developers at strategy center
- 17:44Measuring AI impact
- 20:51Business improvements from AI
- 21:53Lessons learned and next steps
related talks
29mBeyond access: Building the GenAI platform behind 3000+ developers- 29mBeyond access: Building the GenAI platform behind 3000+ developers
40mMaximizing impact: Measuring the ROI of GenAI in the SDLC
23mUnleashing the Power of Gen AI to Boost Engineering Engagement and Productivity
42m10x Engineering: How GenAI Tools Transformed My Team's Productivity
31mAgentic Coding at AirBnb
27mSoftware engineering with GenAI
22mHow Can Engineering Leaders Measure the Impact of GenAI