Store and query high-dimensional vectors with exact k-NN search in Elixir.
npx skills add https://github.com/8dazo/elix-db --skill vector-db-elixir
Please help me install this Agent Skill. Skill: vector-db-elixir Source: https://github.com/8dazo/elix-db/tree/main/.cursor/skills/vector-db-elixir Command: npx skills add https://github.com/8dazo/elix-db --skill vector-db-elixir
Vector databases enable fast similarity search over embeddings within Elixir applications, eliminating manual matching and enabling real-time recommendations and semantic search.
Instantiate an in-memory vector store, upsert vectors, and run a k-NN search to retrieve the most similar vectors.
High-intent search queries and answers about installing and using this skill.
To run semantic search on embeddings in Elixir, you insert vectors with optional metadata into a store and execute a k-NN query. This skill supports exact k-NN search with configurable distance metrics like cosine, L2, and dot product to find similar vectors.
For k-NN search in an Elixir vector database, you can use cosine similarity, L2 (Euclidean distance), and dot product metrics. These configurable distance metrics allow you to fine-tune similarity matching for different embedding models and recommendation use cases.
Yes, you can persist vector embeddings using pluggable backends in Elixir. This skill supports ETS for in-memory storage, file persistence for local saving, and external databases, allowing you to choose the right storage mechanism for your application scale.
Yes, this Elixir vector storage skill includes an optional HTTP API for RESTful access. This allows external services to insert, get, delete, and search vectors over the network, making it easier to integrate similarity search into broader application architectures.
Exact k-NN search provides precise similarity matching for recommendation systems in Elixir, but it may face performance limits at very large scales. For massive datasets, you might need approximate nearest neighbor solutions, though exact k-NN ensures perfect accuracy for stored embeddings.
ArdurAISelect vector databases and train embedding models for approximate nearest neighbor search.
MesferAliPerform vector similarity search and storage using the Qdrant database.
zjunlpParses Alfworld observation text to catalog visible objects and receptacles.
twjarviszyra-webBuild exact and approximate FAISS similarity search indexes for dense vectors.
