Full conference ticket
Full access to all conference sessions, keynotes, and networking events.
In person · Tuesday, May 18 to Wednesday, May 19, 2027
- Meals included
- Swag included
Events /
May 17 – 19, 2027 · San Francisco, CA
Full access to all conference sessions, keynotes, and networking events.
In person · Tuesday, May 18 to Wednesday, May 19, 2027
Workshops run for half a day, in the morning or the afternoon.
New to Temporal? Start here.
You'll learn the core model (Workflows, Activities, Workers and the Temporal Service, including Event Replay) and why durable execution changes how you write reliable code. Each concept is taught once, shown in Python, and then practiced hands-on in the SDK you use at work.
By the end you'll have built and run a real Temporal application, then killed its Worker mid-run and watched it pick up exactly where it left off. Language-specific gotchas and exceptions will be pointed out in the course of the day.
In this workshop you'll build an AI agent on Temporal so every model call, tool call and human approval survives crashes and restarts. You'll run tool calls as durable Activities, absorb rate limits and flaky model responses with retry policies, pause for human review without losing state, and trace every decision the agent made in the event history.
You'll leave with a working agent and the patterns to take it to production.
Workflows can live for days or months, so how do you change their code without breaking them? This workshop answers that in two parts that tell one story.
First, you'll crash a Worker, read the event history to see exactly what happened, and replay a production failure locally in a debugger. Then you'll change one line and hit a non-determinism error.
The second session opens on that broken replay and fixes it, using versioning to ship new code safely alongside the Workflows already in flight, and replay testing to catch the problem in CI next time.
As Temporal spreads across an organization, teams need to call each other's Workflows without sharing a namespace, a codebase or each other's internals. Nexus gives them a contract to do that durably.
In this workshop you'll define a Nexus service, implement its handlers in one namespace, and call it from a Workflow in another, with retries, cancellation and failure handling working across the boundary. You'll leave knowing how to design service contracts that let teams move independently.
You've built your first Workflow; now make it ready for production.
This workshop covers the patterns teams reach for once real traffic arrives: sending data into running Workflows and reading their state with Signals, Queries and Updates; durable timers; Child Workflows, Parallel Activities, and Heartbeats; keeping long-running Workflows healthy with Continue-As-New; and testing Workflows before they ship.
Every pattern is practiced hands-on, so you leave with code you can adapt, not just slides. Language-specific gotchas and exceptions will be pointed out in the course of the day.
In this workshop you'll build an AI agent on Temporal so every model call, tool call and human approval survives crashes and restarts. You'll run tool calls as durable Activities, absorb rate limits and flaky model responses with retry policies, pause for human review without losing state, and trace every decision the agent made in the event history.
You'll leave with a working agent and the patterns to take it to production.
Workflows can live for days or months, so how do you change their code without breaking them?
This workshop answers that in two parts that tell one story.
First, you'll crash a Worker, read the event history to see exactly what happened, and replay a production failure locally in a debugger. Then you'll change one line and hit a non-determinism error.
The second session opens on that broken replay and fixes it, using versioning to ship new code safely alongside the Workflows already in flight, and replay testing to catch the problem in CI next time.
Coding agents made writing code cheap, but we’re still stuck reviewing, approving and shipping it–those are the new bottlenecks.
In the first half you’ll build a delivery pipeline on Temporal that survives real failures: retries, heartbeats on long builds, environment locks, and rollback that undoes a bad deploy. In the second half you’ll put an AI agent in the developer’s seat and decide how far to trust it.
You’ll build a generate-test-fix loop with retry budgets and cost ceilings, approval gates that send each change to a human, a bot or straight to merge based on risk, and feature flags that make merging safer. The event history is the audit trail of everything the agent did, and the first place you look when something breaks. Built for platform and developer-experience engineers.
Your basket is empty.
Discounted prices are an estimate. We confirm the final amount before you pay.