Skip to main content
Aerospike Vector is now live with LangChain integrationTell me more
Loading

Example apps

Overview​

Aerospike Vector Search (AVS) delivers an Approximate Nearest Neighbor (ANN) search using the Hierarchical Navigable Small World (HNSW) algorithm. AVS provides a new set of capabilities and APIs for performing vector operations. The following example apps are available:

Reference applications​

The following provide ideal references for experimenting with AVS and building your own applications:

  • Basic search: A simple Python application that demonstrates vector ANN index creation, vector record insertion, and basic ANN query against an AVS server using the Python client.

  • Quote semantic search: Provides semantic search for an included dataset of quotes by indexing them using the MiniLM model. This model generates vectors with semantic meaning from each quote and stores them as vector embeddings in Aerospike. When a user performs a query, AVS generates a vector embedding for the provided text and performs an ANN search to find relevant results.

  • LangChain Python Notebook: This Jupyter Notebook provides a step-by-step walkthrough of using the native AVS Python client to build the semantic search application for searching quotes that the above example builds out.

  • Prism image search: Provides semantic search for a set of images by indexing them using the CLIP model created by OpenAI. This model generates vectors with semantic meaning from each image and stores it as a vector embedding in Aerospike. When a user performs a query, AVS generates a vector embedding for the provided text and performs an ANN search to find relevant results.

  • Hybrid search: This example app showcases how we built aerospike.com search. It uses an inverted index leveraging spAcy as well as a vector index, and ranks the results.

Demonstration applications​

The following are used for Aerospike demonstrations. These can be used as reference, but have more specific implementations that may not be applicable to your use case.

  • Aerospike RAG: This retrieval augmentation application is designed to scrape technical documentation, and create a chat-bot to answer your questions about Aerospike (or other technical documentation.)

  • Retail Demo: This demo combines Aerospike Key Value, Graph, and Vector into a comprehensive application leveraging the Kaggle Fashion data set.