Workflow Orchestration, Evolved
Durable execution is emerging as the foundation for reliable AI workflows and durable agents at AI scale. On episode three of the Temporal Video Series, Temporal’s Preeti Somal and Tim Hughes join Alan Shimel to explain why Temporal was ready for this moment.
About Preeti Somal and Tim Hughes
Preeti leads the Build, Operate and Secure organization at Temporal, and Tim is SVP of Sales and Customer Success. Consequently, they bring both the engineering and go to market view of how enterprises run mission critical workflows.
Inside durable execution
Durable execution is the programming model that lets a workflow survive crashes, restarts, retries and long running dependencies. As a result, durable agents can pick up where they left off instead of losing state when a browser tab closes or a third party API times out. Meanwhile, the same durable execution pattern already powers every Snap story and every Uber ride, so it maps cleanly onto agentic AI.
In addition, Preeti and Tim discuss AI scale, governance and audit trails, the Andrej Karpathy point that you do not want an LLM guessing a retry policy, and the pattern of hollowing the core when you modernize a brownfield stack. Therefore, teams get durable execution as a small first step and expand from there.
Why this matters now
Meanwhile, enterprises are moving from AI experiments to production, so reliability, visibility and open source foundations are non negotiable. Consequently, durable execution becomes the control plane for agentic workflows that have to be up, observable and scalable under real business pressure.
Explore more artificial intelligence coverage and the latest Techstrong TV interviews. Furthermore, catch episode four of the Temporal Video Series soon.
For more information please visit temporal.io
