Episode

Maxim Fateev - Temporal

Podcast: devtools.fm: Developer Tools, Open Source, Software Development
Published: Jun 23, 2025
Duration seconds: 3129
Processing state: processed
Canonical source: https://podcasters.spotify.com/pod/show/devtoolsfm/episodes/Maxim-Fateev---Temporal-e34jcbh
Audio: https://anchor.fm/s/dd6922b4/podcast/play/104492849/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2025-5-24%2F402704185-44100-2-61d26f26954a5.m4a
JSON: /v1/public/podcasts/devtools-fm/episodes/maxim-fateev-temporal
Markdown: /podcast/devtools-fm/maxim-fateev-temporal.md

Actions

POST https://stenobird.com/v1/public/podcasts/devtools-fm/episodes/maxim-fateev-temporal/transcription-requests
Idempotently request low-priority transcript generation for this episode.
GET https://stenobird.com/podcast/devtools-fm/maxim-fateev-temporal.md
Read the agent-friendly Markdown representation of this episode resource.

Summary

Maxim Fateev, co-founder of Temporal, explains how 'durable execution' abstracts the complexities of distributed systems. He details how Temporal provides a reliable foundation for long-running workflows, from traditional business logic to modern AI agents.

Topics

Durable Execution
Distributed Systems
Temporal
AI Agents
Software Architecture
Event Sourcing
Workflow Orchestration
Open Source Software

Highlights

Main idea: Durable execution abstracts distribution, allowing developers to write standard code that survives machine failures and network partitions
Practical takeaway: Use Temporal to manage stateful, long-running processes like customer onboarding or AI agent orchestration without manual retry logic
Failure mode: Traditional background tasks and queues require developers to manually reassemble complex patterns, leading to inconsistent and fragile implementations
Technical insight: Temporal uses event sourcing to capture side effects, ensuring that even if a process runs for months, its state remains recoverable
Future trend: The rise of AI agents necessitates reliable, long-running tool calls, positioning durable execution as a critical layer for the agentic ecosystem

Chapters

1:00 Background and Engineering Roots: Maxim discusses his experience at Amazon, Microsoft, Google, and Uber, and his transition from IC to CTO.
5:00 The Concept of Durable Execution: An introduction to the new abstraction of durable execution and why traditional distributed patterns fail to scale.
8:50 Handling Failures and Registration Logic: How Temporal removes the need for developers to write complex registration and error-handling logic.
12:45 Language Agnostic Workflows: Exploring how Temporal supports various languages and allows for standard programming constructs like async/await.
16:50 The Architecture of the Temporal Service: A deep dive into the task queues, state updates, and the underlying service mechanics.
20:55 Managing Large Payloads and State: Technical strategies for handling large data transfers and managing state machines in the backend.
24:50 Orchestrating AI Agents and Business Flows: How Temporal serves as the backbone for complex business processes and the emerging field of AI agents.
32:35 Versioning and Long-Running Processes: How to manage code updates and bugs in processes that are designed to run for weeks or months.