The durable execution company teaching software a simple trick: how to survive its own failures.
THE MARK. Temporal's brand sits on the least visible layer of modern software - the part that quietly keeps running after everything else falls over. Company logo, temporal.io
Every engineer who has shipped a long-running process knows the quiet dread: a server reboots, a network blips, an API times out, and somewhere a half-finished job is lost with no clean way to recover. Temporal Technologies built a company around removing that dread. Its durable execution platform records the full running state of an application - every step, retry, timer and result - so that when something breaks, the code simply picks up where it left off.
The pitch the engineering team uses is deliberately plain: write code as if failures don't exist. In practice, developers write ordinary functions - called Workflows and Activities - and Temporal handles the machinery underneath. It persists state, retries failed steps with configurable backoff, coordinates timers and signals, and replays a workflow's history to reconstruct exactly where it was after a crash. What used to require a tangle of queues, cron jobs, dead-letter handling and bespoke state tables collapses into a single, coherent programming model.
That model matters because the alternative is expensive in ways that don't show up until production. Distributed systems fail in dull, relentless ways, and most of the code teams write to survive those failures is glue - fragile, hard to test, and easy to get subtly wrong. Temporal absorbs that glue into infrastructure, which is why it tends to spread inside a company: one team adopts it for a payment flow, and soon the media pipeline, the onboarding system and the provisioning jobs are running on it too.
Production applications require a certain level of resiliency, scale, and observability.
Temporal's founders did not stumble into this problem. Maxim Fateev and Samar Abbas met building large-scale systems at companies like Amazon, Microsoft, Google and Uber. At AWS, they helped design Simple Workflow, one of the first services to tackle application state management head-on. At Uber, they created Cadence, an open-source orchestration engine that became the seed of everything that followed.
In 2019 they left to build the idea the way they always wanted it - open source at the core, developer-first, and unencumbered by a big company's constraints. That project became Temporal. The lineage shows: this is durable execution designed by people who had already shipped it twice and knew exactly what they wanted to fix. In 2024, Fateev handed the CEO role to co-founder Samar Abbas and moved to CTO, a quiet handoff that kept the company focused as it scaled.
The open-source durable execution engine. Self-hostable, free, and the foundation the whole company is built on.
A fully managed, consumption-based version of the service - billed on actions, active workflows and storage - so teams skip the operational burden.
Write workflows as code in Go, Java, TypeScript, Python, .NET, PHP and Ruby. Meet developers in the language they already use.
Patterns and packaging for building reliable, stateful, long-running AI agents on top of durable execution.
A serverless durable execution option unveiled at Replay 2026, lowering the barrier to getting started.
Temporal's annual developer gathering - named for how the engine replays workflow history to recover state.
Temporal is the kind of infrastructure most people never see and many companies can't do without. More than 3,000 paying customers and hundreds of thousands of developers run on it, processing billions of transactions a day.
At Netflix, Temporal powers reliable multi-cloud operations - including Spinnaker, the platform behind the vast majority of Netflix's software deployments - and helped cut transient deployment failures from roughly 4% to 0.0001%. At Snap, Temporal Cloud provides a reliable, scalable foundation for consumer features such as Snap Stories. The through-line: wherever a workload has to run for a long time, hold state, and never quietly disappear, Temporal tends to be underneath it.
Temporal has raised roughly $754M since 2019. The February 2026 Series D - led by Andreessen Horowitz, with Lightspeed, Sapphire and insiders including Sequoia, Index and Tiger - doubled the company's valuation to $5B, funding a push to make durable execution the reliability layer for agentic AI.
Temporal follows an open-core playbook. The engine is free and open source, which drives adoption and trust among developers; revenue comes from Temporal Cloud, the managed service that removes the operational pain of running the durable execution backend at scale. Cloud pricing is consumption-based - metered on actions, active workflows and storage retention - with tiers from a modest monthly minimum up to custom enterprise agreements that add SSO, higher limits and provisioned capacity.
| Approach | What it optimizes for | Trade-off vs. Temporal |
|---|---|---|
| AWS Step Functions | Managed, AWS-native orchestration | Config/JSON-first and cloud-locked vs. code-first and portable |
| Apache Airflow | Batch data pipelines / scheduling | DAG scheduler, not general durable app logic |
| Camunda | BPMN business process modeling | Model-driven vs. write-real-code durable execution |
| Prefect / Dagster | Data orchestration | Data-workflow focus vs. general microservice reliability |
| In-house engines | Full control | You build and maintain the hard reliability layer yourself |
Temporal's differentiator is the combination: an open-source, code-first model that works across many languages, applied to general application logic rather than just data pipelines or business-process diagrams - all backed by a managed cloud.
Maxim Fateev and Samar Abbas launch Temporal, building on their open-source Cadence work from Uber.
An $18.75M Series A funds early development as the open-source community grows.
An Index-led round pushes valuation past $1.5B; Temporal Cloud reaches general availability.
A B-Prime extension keeps the managed platform scaling.
A flat-valuation round funds a push into durable execution for agentic AI, plus a $105M secondary.
Andreessen Horowitz leads a round that doubles the valuation and funds the agentic-AI reliability bet.
Temporal builds a durable execution platform that automatically saves application state and retries and resumes code across failures, letting developers write reliable long-running workflows as if crashes never happen.
Yes. The core Temporal engine is open source and free to self-host. Temporal also sells Temporal Cloud, a fully managed version, plus enterprise support.
More than 3,000 paying customers and hundreds of thousands of developers, including Netflix, Snap, Stripe and Nvidia, along with many AI projects and labs.
About $754M in total, including a $300M Series D in February 2026 led by Andreessen Horowitz at a $5 billion valuation.
It was co-founded in 2019 by Maxim Fateev and Samar Abbas. Abbas is now CEO and Fateev serves as CTO.