Shopify's 'Flywheel': a continual learning loop on PyTorch + vLLM that beats frontier models at a fraction of the cost
PyTorch · x · 2026-11-10
At PyTorch Conference North America, Shopify's Anthony Mazza presented "Flywheel," a state-of-the-art continual learning loop built on PyTorch and vLLM that Shopify says outperforms frontier models at a fraction of the cost.
The core insight: as enterprises scale generative AI, the bottleneck shifts from building an initial model to maintaining and improving performance in production. Shopify turns real-world failures into a self-improving engine rather than letting a static model degrade after launch. Full keynote video is on YouTube.
More from Companies & People
- Before Rushing to Judge Jensen Huang: Investigation Should Follow Evidence, Not Leaks — pstAsiatech · 2026-11-10
- Three career lessons for data engineers: business outcomes, tech debt, soft skills — Zachly · 2026-11-10
- Cohere's Aidan Gomez slams AI doomers: their safety fix is always less competition — cohere · 2026-11-10
- Anthropic's Fable 5.2 costs 2.6× Opus 5.5 per task—and Anthropic says start with Opus — demian_ai · 2026-11-10
- 'The Day He Woke Up a Loser': X users riff on a hypothetical Jensen Huang movie — JacquesThibs · 2026-11-10
- A One-Person AI Company in Hong Kong Staffed by a Team of Grok Bots — AIFlow_ML · 2026-11-10