Airbnb CTO Ahmad Al-Dahle details inside-out AI strategy with 60% AI-authored code
Airbnb CTO Ahmad Al-Dahle, formerly head of generative AI at Meta, described the company's "inside-out AI" approach in an interview with Latent Space. The strategy uses AI internally to accelerate product development, then applies those capabilities to transform the guest experience. Al-Dahle said 60% of Airbnb's code is now AI-authored, feature shipping has increased nearly 80% year over year,…
Key points
- 60% of Airbnb code is AI-authored, feature shipping up nearly 80% YoY, PR throughput up 1.6x
- Internal context graph Everest enabled grocery delivery in 8-9 months and airport pickups in ~6 weeks
- Roughly half of support tickets resolved by AI; at least 10 customized models deployed in production
The company moved from traditional handoffs between product, design, and engineering teams to working directly with prototypes, making code the primary artifact. An internal context graph called Everest, built with LLMs and embeddings, helped launch grocery delivery in eight to nine months and airport pickups in about six weeks. Roughly half of customer support tickets are now resolved by AI, with synthetic data used for pre-production testing. Airbnb deploys at least 10 customized models, selecting from frontier and open models based on cost, performance, and latency trade-offs for each use case. The company is also implementing asynchronous agents for tasks like automated on-call triage.
The story so far
2 episodes →- Airbnb CTO Ahmad Al-Dahle details inside-out AI strategy with 60% AI-authored codethis story
Inside-Out AI: Rebuilding Airbnb Behind the Scenes and Across the Guest Experience
Latent Space · 2 October 2026
Loading the full article…
This text was published by Latent Space and written by Richard MacManus. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Enterprise & Industry
All →- Enterprises struggle to operate AI agents beyond routine tasks · 1 src
- Clay’s Kareem Amin discusses GTM engineer role at Disrupt 2026 · 1 src
- Microsoft warns AI agents expand cyberthreats across enterprise systems · 4 src
- Forward deployed engineers see 729% job growth but face burnout risks · 1 src
- Microsoft launches three real-time voice AI models for developers · 5 src
Comments
via GitHub Discussions