Presentation: Dynamic Moments: Weaving LLMs into Deep Personalization at DoorDash
📰 InfoQ AI/ML
Learn how DoorDash uses LLMs for dynamic personalization, generating consumer profiles and content blueprints, and combining with traditional deep learning for last-mile ranking
Action Steps
- Build a hybrid model using LLMs and traditional deep learning
- Generate natural-language consumer profiles using LLMs
- Create content blueprints using LLMs
- Implement last-mile ranking using traditional deep learning
- Integrate the hybrid model into a production-ready pipeline
Who Needs to Know This
Data scientists and engineers at companies like DoorDash can benefit from this approach to improve personalization and user experience, while product managers can use this to inform product strategy
Key Insight
💡 Hybrid approach combining LLMs and traditional deep learning enables dynamic, moment-aware personalization
Share This
🚪 DoorDash's dynamic personalization uses LLMs to generate consumer profiles & content blueprints, combined with traditional deep learning for last-mile ranking #LLMs #Personalization
DeepCamp AI