TechPrufer
Sign inGet started

Blogs›Modal›Hybrid Search with Embeddings

Hybrid Search with Embeddings

Hybrid Search with Embeddings

1
posts
2024

Modal now supports building hybrid search applications by integrating with data platforms like MongoDB Atlas and foundation models for generating embeddings. This enables complex search queries combining text, geolocation, timestamps, and vector embeddings derived from satellite imagery, demonstrating a new capability for data-driven AI applications.

2024

Hybrid search over California embeddings with Modal, MongoDB, and Clay

9/24/2024

This post details the architecture and implementation of a hybrid search system using Modal, MongoDB Atlas, and Clay's foundation model for satellite imagery embeddings. It covers data ingestion from Sentinel satellites, embedding generation on Modal GPUs, storage and querying in MongoDB Atlas, and a frontend for hybrid search requests. The technical depth includes specific code paths for data extraction, embedding generation, and database interaction, showcasing the integration of diverse services for a complex AI application.