Temporal founders: Samar Abbas and Maxim Fateev
Temporal founders Maxim Fateev and Samar Abbas explain how durable execution lets teams write business logic as if common failures do not exist — and how open source, consumption-aligned sales and years of credibility earn enterprise trust.
Maxim Fateev and Samar Abbas co-founded Temporal after building workflow platforms across Amazon, Microsoft and Uber. Temporal persists the state of long-running processes so applications can recover from infrastructure failures and still complete.
We talk about processing every Snapchat story, building once instead of rewriting after product-market fit, forking their earlier open-source work, recruiting outside a big-tech network, landing critical enterprise workloads and paying sales teams when customers actually consume the product.
The Q&A below is experimental — extracted from the episode transcript with AI. If you spot a mistake, let us know at jack@scalingdevtools.com.
What does Temporal do for Snapchat stories?
Each uploaded story triggers a workflow that orchestrates multiple backend actions before it reaches other timelines. Temporal ensures those calls complete even during extreme peaks such as New Year's Eve. It sits in the processing path of every story, making reliability part of a very visible consumer experience.
Why use Temporal at the earliest startup stage?
Startups optimize for time to market, then often replace the first architecture once growth demands reliability and scale. Temporal is faster at the beginning because developers write business logic without hand-building retries and recovery. The same code can then scale, avoiding a rewrite that pulls engineering away from customer value.
What does “write code as if failures don’t exist” mean?
The platform persists workflow state, queues, timers and execution history, then resumes work through infrastructure failures. Developers still design the business process, but many common failure-handling concerns disappear from application code. It is a programming-model change, not simply a retry library.
How did your careers lead to Temporal?
Both of us spent years building platforms at very large technology companies and first worked together on Amazon Simple Workflow. We reunited at Uber, where the predecessor technology ran hundreds of production use cases. Temporal became the next iteration of a problem we had been refining for roughly two decades.
Why fork the earlier open-source project?
The MIT license allowed it, while a fork gave us one chance to make changes that could not remain backward compatible. We then spent about a year building before blessing the first production release. After that point, protecting compatibility again became a core promise to customers.
What changes when founders recruit beyond their network?
Clarity becomes essential: explain the founding story, the size of the problem and the impact a new person can have. Prestige and compensation at a big company make departure genuinely risky. Early proof points, market momentum and visible technical credibility gradually make that leap easier for candidates to justify.
How does open source help earn enterprise trust?
Engineers can try the technology, assess its developer experience and even run the backend themselves before committing commercially. That broad access combines with the founders' production history and an incremental adoption path: begin with one important workflow, prove reliability, then expand rather than demanding a platform-wide migration.
Why compensate salespeople based on consumption?
A contract alone is not success. Consumption arrives only after the customer has identified a real workload, built it successfully and put it into production. Tying compensation to that lagging signal aligns sales with customer outcomes and discourages deals that become unused shelfware.
What makes Temporal sales technical?
Account executives work with solutions architects, many of whom come directly from development backgrounds. New-logo hunters and existing-account teams share the same requirement: understand the workload, solve a real problem and help it reach production. Engineers and even the founders remain visible in community support as well.
Where do you see AI changing development platforms?
Today's assistants often feel like very capable autocomplete, but the broader direction resembles a new low-code movement. Generated applications still need dependable execution underneath. Temporal already powers low-code systems, so the same reliability layer can let AI-created workflows run at scale without every creator becoming a distributed-systems expert.