Inside Pinecone Nexus: A knowledge layer for agents

Pinecone's VP of Product, Jeff Zhu, and Principal Product Manager, Jasmeet Singh Gujral, sit down with hosts Jenna Pederson and Arjun Patel to introduce Pinecone Nexus, a new knowledge layer built specifically for AI agents. They unpack why traditional RAG and raw agentic search waste tokens and time by forcing every agent to re-derive its own context, and how Nexus's knowledge artifacts and Knowledge Query Language (KnowQL) let agents get accurate, grounded answers without that repeated exploration. The team then walks through a live demo of the Nexus, from ingesting a wealth-advisory dataset to designing a manifest, curating knowledge artifacts, and querying with full trace visibility. They close with real customer benchmarks and answer audience questions on pricing, security and RBAC, and how to get access to the public preview. Resources: 🔗 Pinecone Nexus Public Preview announcement: https://www.pinecone.io/blog/pinecone-nexus-public-preview/ 🔗 Request access to Nexus Public Preview: https://www.pinecone.io/lp/nexus-public-preview/ Follow Jeff: https://www.linkedin.com/in/jeffrey-s-zhu/ Follow Jasmeet: https://www.linkedin.com/in/jsgujral04/ Follow Arjun Patel: https://www.linkedin.com/in/arjunkirtipatel/ Follow Jenna Pederson: https://www.linkedin.com/in/jennapederson/ https://twitter.com/jennapederson https://www.instagram.com/jennanerdsout/ Follow Pinecone: https://twitter.com/pinecone https://www.linkedin.com/company/pinecone-io/ https://discord.gg/qwQPPPZ4nz 0:00 Kickoff 5:29 From vector search to RAG: Pinecone's early journey 9:04 Why agents need a smarter knowledge layer 13:53 Introducing Nexus 17:21 KnowQL: a query language built for agents 21:35 Lowering the barrier to building knowledge layers 27:30 Live demo: importing data, designing a manifest, querying Nexus, and exploring traces 48:05 Customer benchmarks, public preview access, and Q&A #Pinecone #Nexus #AIAgents #AgenticAI #RAG