Presentation: Dynamic Moments: Weaving LLMs into Deep Personalization at DoorDash
DoorDash isn't just using LLMs for chat. They're wiring them into the core ranking engine—adapting menus in real-time based on user intent.
Why it matters
This is a practical case study in how mature AI companies are moving beyond chatbots to embed LLMs into revenue-critical infrastructure. The hybrid approach (LLMs for profiling + deep learning for ranking) offers a scalable blueprint for personalization at scale.
The key facts
10 to knowDoorDash shifted from static to dynamic, moment-aware personalization
LLMs generate natural-language consumer profiles and content blueprints
Traditional deep learning handles last-mile ranking
Hybrid approach adapts to short-lived user intent and massive catalog abundance
Presented by Sudeep Das and Pradeep Muthukrishnan at InfoQ
LLMs generate natural-language consumer profiles at DoorDash
Hybrid approach: LLMs for intent/content, deep learning for ranking
Focus on moment-aware personalization vs. static merchandising
Addresses massive catalog abundance problem
Adaptive to short-lived user intent signals
Go to the source
InfoQ AI/MLinfoq.com
Publisher excerpt: Sudeep Das and Pradeep Muthukrishnan explain the shift from static merchandising to dynamic, moment-aware personalization at DoorDash. They share how LLMs generate natural-language "consumer profiles" and content blueprints, while traditional deep learning handles last-mile ranking. This hybrid…