Subvolts take Secure sandboxes are essential for AI agents, moving beyond current cloud infrastructure. May 14, 2026 - Developers
- Learn about the necessary infrastructure and security for building and deploying AI agents effectively.
- Business
- Understand the technical requirements and emerging solutions for running autonomous AI workloads securely.
This episode provides a deep dive into the practical infrastructure needed for AI agents, a key area for future AI development and deployment.
The Agent Harness: Building Secure Sandboxes for Autonomous AI Workloads
Audio streams from Matt Turck. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
On The MAD Podcast, host Matt Turck interviews Ivan Burazin, CEO of Daytona, about the infrastructure for AI agents. They discuss the need for secure, stateful "computers" or sandboxes for AI agents, contrasting them with legacy stateless hyperscalers. The conversation covers the technical aspects of sandboxes, agent memory, and the emerging AI agent stack, along with product-led growth strategies for technical founders.
Imagine AI agents as digital workers. This episode explains they need their own secure digital workspaces, like a personal computer, to do their jobs safely and effectively. This is different from how most cloud services work today.
This episode delves into the practical infrastructure required to run AI agents, a critical area as companies like OpenAI, Anthropic, and Google Gemini develop more sophisticated agent capabilities. It's for developers and technical leaders building or deploying AI agents, explaining why current cloud solutions are insufficient. Understanding sandboxes and the agent stack is key to managing the security and statefulness of autonomous AI workloads, directly impacting how frontier AI models will be utilized in production.
- HostMatt Turck
- GuestIvan BurazinCEO, Daytona
- Guest: Ivan Burazin, CEO of Daytona
- Show: The MAD Podcast
- Length: 65 minutes
- Topic: Emerging infrastructure stack for AI agents and secure sandboxes
- Key Concept: Every agent needs its own secure, stateful "computer" (sandbox)
If AI agents are the new digital knowledge workers, where exactly do they do their work? In this episode of the MAD Podcast, Ivan Burazin, CEO of Daytona, joins us to unpack the emerging infrastructure stack for AI agents and explain why every agent needs its own secure, stateful "computer." We explore the technical realities of sandboxes, dive into why legacy, stateless hyperscalers weren't built for these new workloads, and break down the mechanics of microVMs and custom schedulers alongside a contrarian prediction on an impending CPU shortage. Finally, Ivan delivers an absolute masterclass on product-led growth, community building, and go-to-market strategy for technical founders.
(00:40) Intro
(02:13) What is an AI agent sandbox?
(03:17) Security risks of running agents locally
(05:17) Stateful vs. stateless hyperscalers
(07:04) The history of cloud IDEs and the end of localhost
(09:45) Do all AI agents need a sandbox?
(12:26) Sandbox use cases: RL evals & background agents
(14:10) Unpacking the emerging AI Agent Stack
(16:20) The unsolved problem of agent memory and learning
(19:37) Where sandboxes fit in the agent harness
(21:35) OpenAI, Anthropic, and agent SDKs
(23:06) Ivan's founder journey: From CodeAnywhere to Daytona
(26:59) GTM strategies and building developer communities
(33:48) Why customer support is your best GTM strategy
(35:34) Leveraging Twitter during the AI super cycle
(40:50) The technical anatomy of a sandbox
(41:53) Why fast spin-up speeds maximize GPU efficiency
(46:09) Firecracker, QEMU, and isolation primitives
(49:58) Why sandbox snapshots and state forking matter
(51:40) Why Daytona built a custom scheduler from scratch
(55:24) The challenge of long-running stateful sandboxes
(58:10) The build your own sandbox trap
(1:01:03) Why AI agents might trigger a global CPU shortage
(1:02:46) The future of the AI Agent Stack
- Stateful vs. stateless hyperscalers
- Ivan Burazin
- Cloud computing services that either remember past interactions (stateful) or start fresh each time (stateless).
- MicroVMs
- Ivan Burazin
- Very small, efficient virtual machines used to isolate and run applications, often for security and speed.
34 minMusk v Altman: Much ado about nothing
WithLiz Lopatto
52 minHermes Agent: Agents that grow with you
WithJeffrey Quesnelle
31 min
1 h 21Google Changes Search for the First Time in 25 Years
WithAndrej Karpathy
Questions people ask
- What is an AI agent sandbox?
- An AI agent sandbox is a secure, stateful environment where AI agents perform their work, acting like a dedicated computer for each agent.
- Why do AI agents need sandboxes?
- Sandboxes provide necessary security and statefulness for AI agents, addressing limitations of legacy, stateless cloud infrastructure not designed for these new workloads.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Matt Turck.