18 Months of Pgvector Learnings in 47 Minutes (Tutorial)

Avthar Sewrathan shares best practices, common pitfalls, and tools for building AI applications using pgvector and PostgreSQL.

​​PostgreSQL is catching fire as the default vector database choice for AI applications. But how can you best leverage all that PostgreSQL has to offer for AI applications? What are best practices to follow, common pitfalls to avoid, and tools that will accelerate your development? ​​​ Avthar Sewrathan, shares his learning from helping developers build AI applications with PostgreSQL over the past 18 months. Avthar covers the state of the union of developing AI applications in 2024, and why PostgreSQL can save you time and headaches now and in the future. 🛠 𝗥𝗲𝗹𝗲𝘃𝗮𝗻𝘁 𝗥𝗲𝘀𝗼𝘂𝗿𝗰𝗲𝘀 📌 Github with all Code, Slides, Resources ⇒ tsdb.co/busy-dev-resources 📌 Free Trial of Tiger Cloud ⇒ tsdb.co/busy-dev-signup 📌 pgai ⇒ tsdb.co/pgai 📌 pgvectorscale ⇒ tsdb.co/pgvectorscale 🐯 𝗔𝗯𝗼𝘂𝘁 Tiger Data At Tiger Data, we see a world made better via innovative technologies, and we are dedicated to serving software developers and businesses worldwide, enabling them to build the next wave of computing. Tiger Data is a remote-first company with a global workforce backed by top-tier investors with a track record of success in the industry. 💻 𝗙𝗶𝗻𝗱 𝗨𝘀 𝗢𝗻𝗹𝗶𝗻𝗲! 🔍 Website ⇒ tsdb.co/homepage 🔍 Slack ⇒ slack.timescale.com 🔍 GitHub ⇒ github.com/timescale 🔍 Twitter ⇒ twitter.com/timescaledb 🔍 LinkedIn ⇒ linkedin.com/company/tigerdata/ 🔍 Tiger Data Blog ⇒ tsdb.co/blog 🔍 Tiger Data Documentation ⇒ tsdb.co/docs 📚 𝗖𝗵𝗮𝗽𝘁𝗲𝗿𝘀 00:00 What You’ll Learn Today 02:03 Foundations of AI with Postgres 03:58 Understanding Vector Data and Databases 06:25 Types of AI Applications with Postgres 08:53 Postgres Extensions for AI 14:25 Demo: Using pgVector and pgAI 21:42 Vector Search Indexes in Postgres 25:14 Vector Compression and Indexing Options 26:13 Introducing Streaming Disk Index 28:12 Creating Vector Search Indices in Postgres 30:22 Advanced Topics in AI Systems 32:14 Evaluation Driven Development 35:51 Filtered Vector Search 40:32 Hybrid Search Techniques 42:31 Multi-Tenancy in RAG Applications 44:41 Text to SQL in AI Applications 45:33 Conclusion and Resources