Vector Search on the Edge Powering Mobile and IoT with AI
summary
Shows running vector search entirely on-device using Couchbase Lite with a new beta vector field stored as descriptionVector in JSON and queried via SQL-like similarity. Uses AlexNet 3D embeddings and k-NN: images are embedded, stored, and nearest neighbors return categories. Demo app is a quiz that classifies uploaded images (triangle → shape) locally, syncs back through Sync Gateway, and supports hybrid mode (on-device small models + cloud for heavy inference). Performance gains: lower latency, offline operation, and retained data privacy.
AI-written summary · about
chapters (7)
- 0:00Introduction to AI on Edge
- 3:52What is Vector Search
- 10:39Couchbase and Vector Search
- 13:46Vector Search on Mobile and IoT
- 18:41Demo Architecture Overview
- 21:32Live Demo and Code Walkthrough
- 29:15Q&A and Conclusion
related talks
33mBringing Vector Search to OLTP
57mEnter The Brave New World of GenAI with Vector Search
33mEnter the Brave New World of GenAI with Vector Search
38mAI-powered Vector Search Engine Weaviate
27mUnified AI & Relational Power: Scalable Distributed Vector Search with
44mVector Search Beyond the Hype: Real-World Implementation and Performance Insights
50mVector Ops: run vector embedding-powered apps in prod
24mVector Search at Uber