- Developers
- Understand the infrastructure challenges and solutions for building reliable, long-running AI agents.
- Business
- Learn how robust infrastructure enables scalable and dependable AI agent deployments for your services.
This interview with the CEO of Temporal provides insight into the infrastructure supporting major AI agent deployments.
Durable Execution and the Infrastructure Powering AI Agents
Audio streams from a16z. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
a16z's Raghu Raghuram and Sarah Wang interview Temporal CEO Samar Abbas about durable execution for AI agents. They discuss how state management and recoverability are crucial for long-running agents. The conversation covers Temporal's role in powering systems like OpenAI's Codex and Snap's Story processing, highlighting the growing distributed systems challenges as agents shift to background tasks.
Imagine AI agents as assistants that need to remember tasks and pick up where they left off. This episode explains the technology that makes these AI assistants reliable, even if they run for a long time or encounter problems. It's important because it helps build more dependable AI services we can all use.
This episode explains the critical infrastructure powering AI agents, a key area for all frontier AI labs. Listeners will learn about the technical challenges of managing state and ensuring reliability for AI systems that run continuously. It is for developers and technical leaders building or deploying AI agents, showing how companies like OpenAI and Snap rely on specialized infrastructure to scale their AI applications. This offers insight into the foundational technology enabling the next wave of AI services.
- Hosta16z
- GuestSamar AbbasCEO, Temporal
- Guest: Samar Abbas, CEO of Temporal
- Hosts: Raghu Raghuram and Sarah Wang, a16z
- Show: AI + a16z
- Length: 64 minutes
- Temporal powers: OpenAI's Codex and Snap's Story processing
Raghu Raghuram, Managing Partner at a16z, and Sarah Wang, General Partner at a16z, speak with Samar Abbas, CEO of Temporal, about how durable execution became the infrastructure layer behind some of the world’s most widely used AI agents. They cover why long-running agents require state management and recoverability, how Temporal powers OpenAI’s Codex and Snap’s Story processing, and why the shift from interactive to background agents is creating distributed systems challenges at a scale that didn’t exist two years ago.
Resources:
Follow Samar Abbas: https://x.com/SamarAtTemporal
Follow Sarah Wang: https://x.com/sarahdingwang
Follow Raghu Raghuram: https://x.com/RaghuRaghuram
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.
- durable execution
- Samar Abbas
- Technology that ensures AI agents reliably manage their tasks and remember progress, even if they stop and restart.
- state management
- Samar Abbas
- Keeping track of all the information an AI agent needs to know to continue its work correctly.
How AI Is Rewriting the Rules of Cybersecurity | Truffle Security & Socket
WithDylan Ayrey, Feross Aboukhadijeh
Inside vLLM: The Engine Powering Open-Source AI
WithSimon Mo
Building Self-Accelerating AI with Mirendil
WithBehnam Neyshabur, Harsh Mehta
Ideogram’s Open-Weights Image Model and the Future of AI Design
WithMohammad Norouzi
Can AI Replace Teachers? Inside the $40M Company Using AI Tutors to Teach 200% Faster | #233
WithMackenzie Price, Joe Lamont
Questions people ask
- What is durable execution in the context of AI agents?
- Durable execution refers to the infrastructure layer that enables AI agents to manage their state and recover from failures, ensuring they can run reliably for extended periods.
- Which companies use Temporal for their AI agents?
- The episode mentions that Temporal powers systems like OpenAI's Codex and Snap's Story processing.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to a16z.





