Dynamic Moments: Weaving LLMs into Deep Personalization at DoorDash

The presentation discusses how DoorDash is leveraging cutting-edge large language models (LLMs) and deep learning architectures to redefine personalization. This involves integrating LLMs with existing systems such as Two-Tower Embeddings and Multi-Task Multi-Label models to enhance user experience through dynamic and hyper-personalized content.

Key Points:

• The aim is to capture "dynamic moments," short-lived, high-context events like movie nights or Black Friday, by meeting consumers' evolving needs in real-time.
• A robust product knowledge graph is created using LLMs to automate and structure DoorDash's catalog, enhancing the understanding of both user behavior and product information.
• The synergy between LLMs and deep learning enables real-time dynamic content generation, tailoring the DoorDash app to users' immediate intents and occasions.
• Examples include providing personalized experiences by tailoring product suggestions based on users' recent activities and long-term interests.

Examples:

• Illustration of personalized experience through a user scenario: On Black Friday, a user interested in electronics would see tailored deals relevant to their browsing history, such as noise-canceling headphones.
• Real-time adjustments in content, acknowledging changes in user intent as they navigate the app.

Conclusion:

DoorDash's integration of LLMs and deep learning is transforming consumer engagement by providing instantaneous personalization that anticipates users' needs and adapts to their preferences dynamically.

Overall, the presentation highlights the practical implementation and benefits of machine learning integration for deep personalization.

This is the end of the AI-generated content.

In this talk, we’ll walk through how DoorDash is redefining personalization by tightly integrating cutting-edge large language models (LLMs) with deep learning architectures such as Two-Tower Embeddings (TTE) and Multi-Task Multi-Label (MTML) models.

We'll begin by sharing how delivering truly dynamic, moment-aware experiences first requires a robust, structured understanding of our product catalog. We'll detail how we automated the generation of a product knowledge graph using LLMs, enhanced with fine-tuning and agentic workflows.

From there, we'll demonstrate how the synergy between LLMs and DL-powered personalization unlocks new frontiers in dynamic content generation — powering collections, modules, and real-time app assets designed to win users’ “moments” (e.g., movie nights, flu seasons, study snacks). Our framework turns DoorDash into a moment-ready utility that understands and adapts to users’ evolving needs — becoming a one-stop shop for whatever life throws at them.

Key Takeaways:

1. A practical framework for integrating LLMs with existing deep learning systems like TTE and MTML for real-time personalization.
2. A blueprint for building and maintaining automated product knowledge graphs using fine-tuned LLMs and agentic pipelines.
3. How to personalize app experiences dynamically to win high-intent user moments and occasions.