Subvolts take AI agents need knowledge engines, not just databases, for efficient machine-to-machine interaction. May 5, 2026 - Developers
- Learn how to build more efficient agentic applications by rethinking data storage and reasoning integration.
- Business
- Understand how new AI infrastructure can improve accuracy, performance, and reduce costs for automated systems.
This episode features the CEO of a key AI infrastructure company discussing a new product and a fundamental shift in AI architecture.
From Vector Databases to Knowledge Engines: The Next Layer of AI
Audio streams from a16z. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
Peter Levine interviews Ash Ashutosh, CEO of Pinecone, about the launch of Nexus. They discuss the shift from vector databases to knowledge engines as AI agents become primary software users. The conversation covers why traditional retrieval systems fail for machine-to-machine interactions and how moving reasoning closer to data improves performance, accuracy, and cost.
Imagine AI programs acting like people, using software. This episode explains why current ways of finding information for these AI agents are slow. It introduces a new idea where AI can think about information right where it's stored, making everything faster and cheaper.
This episode is for developers and product leaders building AI agentic applications. It explains the limitations of current retrieval systems for machine-to-machine communication and introduces Pinecone's new approach with Nexus. Listeners learn how optimizing agent workflows by integrating reasoning with data storage can significantly boost efficiency and reduce costs, a critical consideration as frontier labs like OpenAI and Google Gemini push agent capabilities.
- Hosta16z
- GuestAsh AshutoshCEO, Pinecone
- Guest: Ash Ashutosh, CEO of Pinecone
- Title: From Vector Databases to Knowledge Engines: The Next Layer of AI
- Show: AI + a16z
- Length: 46 minutes
- Product Launch: Pinecone Nexus
Peter Levine speaks with Ash Ashutosh, CEO of Pinecone, about the launch of Nexus and the shift from vector databases to knowledge engines. As agents become the primary users of software, they discuss why traditional retrieval systems break down and how AI systems need to evolve to support machine-to-machine interactions.
The conversation explores how agents currently spend most of their time retrieving and reasoning over data, why that approach is inefficient, and how moving reasoning closer to the data can dramatically improve performance, accuracy, and cost. Ash also explains how Pinecone is rethinking the stack for agentic applications, introducing new abstractions, query languages, and developer workflows.
Resources:
Follow Ash Ashutosh on X: https://x.com/ashashutosh
Follow Peter Levine on LinkedIn: https://www.linkedin.com/in/peter-levine-681386172/
Check out everything a16z is doing with artificial intelligence here, including articles, projects, and more podcasts.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
- vector databases
- Ash Ashutosh
- Databases that store information as numerical lists, useful for finding similar items.
- knowledge engines
- Ash Ashutosh
- Advanced systems that not only store information but also understand and reason with it.
- agents
- Ash Ashutosh
- Software programs designed to perform tasks autonomously, often interacting with other software.
- retrieval systems
- Ash Ashutosh
- Systems that find and fetch specific information from a larger dataset.
24 minHow AI Is Rewriting the Rules of Cybersecurity | Truffle Security & Socket
WithDylan Ayrey, Feross Aboukhadijeh
47 minInside vLLM: The Engine Powering Open-Source AI
WithSimon Mo
41 minBuilding Self-Accelerating AI with Mirendil
WithBehnam Neyshabur, Harsh Mehta
43 minIdeogram’s Open-Weights Image Model and the Future of AI Design
WithMohammad Norouzi
22 min
1 hJoanna Stern is not a robot, but she lived with them
WithJoanna Stern
Questions people ask
- Why are traditional retrieval systems insufficient for AI agents?
- Traditional systems break down as agents become the primary users of software, requiring more direct machine-to-machine interactions that current retrieval methods struggle to support efficiently.
- What is the benefit of moving reasoning closer to data?
- Moving reasoning closer to the data can dramatically improve performance, accuracy, and reduce costs for AI systems.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to a16z.