Ecommerce ops · Production

DoorDash uses LLM-generated profiles and content embeddings to improve semantic search and recommendations across verticals

The problem

DoorDash faced a persistent bottleneck where personalization depended on embedding quality, which in turn depended on data quality — but sparse metadata flattened catalog richness across all verticals. Behavioral co-visitation approaches tried to bypass this dependency but could not capture identity, context, and intent.

Workflow diagram · grounded in source
1
LLM profile generation
Ai action
LLMs produce standardized narratives for merchants and items covering ingredients, preparation, attributes, and context.
▾ source quote
“LLMs produce consistent, high-quality narratives for merchants and items such as ingredients, preparation, attributes, or context that reduce reliance on human-labeling efforts”
2
Incremental embedding pipeline
Integration
Metaflow handles incremental re-embedding, only updating entities whose underlying content has changed.
▾ source quote
“We use incremental inference via Metaflow, which only requires re-embedding entities when their underlying content has changed”
3
Embedding index publishing
Output
Computed embeddings are written to persistent storage and index so downstream experiments can consume them consistently.
▾ source quote
“Publishing writes embeddings to persistent storage/index so that downstream experiments can consume them consistently”
4
Semantic query retrieval
Ai action
Queries are embedded online and retrieved against offline store and item profile embeddings so that even rare or novel queries retrieve semantically aligned candidates.
▾ source quote
“Embed the query online and retrieve against offline store/item profile embeddings, so that even rare or novel queries can retrieve semantically aligned candidates”
5
Item-level reranking
Ai action
A fine-tuned Qwen 3 Rerank model scores each candidate using the search query, item profiles, and store profile.
▾ source quote
“a fine-tuned Qwen 3 Rerank model that scores each candidate by consuming the search query, the item profiles of the top-k most relevant items within a store, and the store profile”
6
Generative personalized carousels
Ai action
An LLM generates a carousel theme from the consumer profile and context, then embeds the theme to retrieve nearest-neighbor stores and representative dishes.
▾ source quote
“An LLM generates a carousel theme from the consumer profile and context, such as time of day, then embeds the theme and retrieves nearest-neighbor stores and representative dishes within the delivery radius. Final ordering uses the existing store ranker, optionally …”
Reported outcome

Deploying LLM-generated profiles with content embeddings improved semantic search and homepage discovery: null search rate fell 3.65%, core search CVR rose 0.66%, and generative personalized carousels drove a 2.4% relative increase in homepage order rate.

Reported metrics
7D active customer share+0.0724%
Null search rate−3.65%
core search session CVR+0.66%
Dish query ranking improvement7.8%
Show all 14 reported metrics
7D active customer share+0.0724%
null search rate−3.65%
core search session CVR+0.66%
dish query ranking improvement7.8%
cuisine query ranking improvement1.4%
trial merchant visit rate+0.435%
homepage clicks per impression+0.110%
consumer homepage order rate2.4%
consumer 7-day reorder rate+0.164%
variable profit per order0.32%
offline precision@10 on homepage68% to 85%
item-to-item Hit@5 with LLM profiles+31.22%
item-to-item Hit@5 data and model combined+37.55%
store-to-store Hit@5 data and model combined+209%
Reported stack
LLMsgemini-embedding-001Qwen 3 Rerank modelMetaflowMiniLMGoogle Gemini embeddingsQwen embedding modelstext-embedding-005text-embedding-3-largeOpenAICohere
◆ Does this fit your context?

Compare to your context

Tell us your scale, team, and constraints. We'll show what changes at your size, what fails at your scale, and whether this case is a fit, needs adaptation, or won't scale to you. Free demo, no signup.

Compare to your context →
~30 seconds · free
Source
https://careersatdoordash.com/blog/doordash-llms-to-build-content-embeddings-for-search-and-recommendations/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Deploying LLM-generated profiles with content embeddings improved semantic search and homepage discovery: null search rate fell 3.65%, core search CVR rose 0.66%, and generative personalized carousels drove a 2.4% rel…

What tools did this team use?

LLMs, gemini-embedding-001, Qwen 3 Rerank model, Metaflow, MiniLM, Google Gemini embeddings, Qwen embedding models, text-embedding-005, text-embedding-3-large, OpenAI.

What results were reported?

7D active customer share: +0.0724%; Null search rate: −3.65%; core search session CVR: +0.66%; Dish query ranking improvement: 7.8% (source-reported, not independently verified).

How is this ecommerce ops AI workflow structured?

LLM profile generation → Incremental embedding pipeline → Embedding index publishing → Semantic query retrieval → Item-level reranking → Generative personalized carousels.

Related ecommerce ops cases

WHAT TO DO WITH THIS

Now compare it to your context

This case is one data point. Whether its pattern fits you depends on your volumes, your stack, and your exception load — that comparison is the step no case study can do for you.

Browse the Ecommerce ops cases →