Feature Search
Vector similarity search across embeddings with configurable distance metrics and filtering
Why do anything?
Embeddings are useless without search. You need to find similar vectors to power semantic search.
Why now?
Every AI application needs vector search. Users expect semantic understanding, not keyword matching.
Why this feature?
High-performance vector search with cosine/euclidean/dot distance metrics. Supports pre-filtering and metadata constraints.
How It Works
Feature search is the core retriever stage for vector similarity search across any embedding type.
Query Embedding
Embed query text/image if not already a vector
Vector Search
Search Qdrant with configured distance metric
Filtering
Apply pre-filters on payload fields
Ranking
Return top-k results by similarity
Why This Approach
Qdrant provides efficient ANN search. Pre-filtering reduces search space for better performance.
Where This Is Used
Integration
results = client.retrievers.execute(retriever_id=retriever_id, inputs={"query": "..."})