Beyond The Pilot: Enterprise AI in Action · 2026-05-27 · 34 min
Pinterest's open-source AI stack costs 90% less than frontier models - and their custom-trained recommender outperforms off-the-shelf alternatives by 30% in accuracy. Pinterest CTO Matt Madrigal breaks down exactly how they did it, and what enterprise AI teams can actually replicate. Madrigal walks through the full architecture behind Navigator 1, Pinterest's conversational shopping assistant built on Qwen 3 VL - and the specific decision to rip out its native vision encoder and replace it with PinCLIP, Pinterest's proprietary multimodal embedding layer. That swap alone closes a 20x inference latency gap and makes the economics work at 620 million monthly active users. This is the clearest public explanation yet of how a scaled platform operationalizes the "core vs. context" principle for model selection: open-source and custom-built where it touches the user, frontier models where speed-to-prototype matters more than cost.
Routing Wars: Which Model Gets the Workload?68 / 100
Small Models, Massive Wins: The New Shopify AI Formula85 / 100
Weaponize Tokenmaxing: MassMutual’s ROI Engine
GPU Hoarding is Over. The $401B Reality Check
Agents Ate the UI: Data is Your Only Moat with LlamaIndex