Temporal Technologies

Live this Wednesday, we'll look at what people are actually running with autonomous agents. Melanie Warrick will be joined by Abhay Kashyap, Mark Weiss, and Lynn Conway (Kai Ase Siren), who have spent months building and experimenting with persistent agents. Their setups range from named agents with long-term memories to a collaborative human-agent workspace and a system of eight persistent personas. We'll bring their agents together in a shared Discord for the first time. What can agents from different systems create together? What do they say about their humans when they think nobody is reading? How should permissions work when someone borrows an agent? Watch live and bring questions!

Most infrastructure platforms are brownfield systems. Jenkins jobs, Ansible playbooks, Terraform modules, and delivery pipelines all have owners, history, and interfaces people rely on. Zscaler’s Kartick Krishnachetty Ravi and Abhishek Mitra write about taking a different approach: keep the engines, and put a durable orchestration contract above them with Temporal. That makes it possible to coordinate cross-engine work across ArgoCD, Jenkins, Ansible, and Crossplane without turning another custom pipeline into the most fragile part of the platform. The underlying tools stay intact, including the break-glass paths operators already know. Read the blog: https://lnkd.in/dWRUjaE9

Cool answer. Cooler hair.

Most teams building agents have moved past "can we build a demo" to "can we run this in production." We built a joint reference architecture with MongoDB for exactly that: Temporal for durable execution, MongoDB Atlas and Voyage AI for the data and retrieval agents need to work with. Read the blog for the full breakdown and use cases: https://lnkd.in/dsj87-At If you're at BuildFest in San Francisco today, visit our booth and we'll walk you through it live.

Agents don't need your password. They need your permission. On the next Vibe Check, Jessica Temporal, Senior Developer Advocate at Auth0, joins Melanie Warrick and W. Ian Douglas to dig into how agents get scoped, revocable access to real systems, instead of the usual move of handing them your credentials and hoping nothing goes wrong. Jessica walks through an agent that acts on someone's behalf without ever touching their password, with a human stepping in to approve or deny an action on the spot. Then the harder question most agent demos skip: what happens to that access, and the agent's state, when it has to sit and wait on a human decision, or come back from a crash.

In this edition: expert-reviewed Validated Patterns for production architectures, answers to real questions from the Temporal community, and opportunities to learn directly from the team through upcoming events.

Today’s special at Melissa’s Cafe: agent harnesses. What’s your order?

A human is an async API. Terrible latency, no SLA, and sometimes it returns “ask someone else instead.” Ziggy’s Ice Cream shows how to make that wait durable across Google ADK, LangGraph, and Temporal. Read the blog: https://lnkd.in/dZEB9yie

At our annual conference, Replay, we had two nights and one workday to flash, assemble, and charge 2,300 conference badges before they hit the stage. We used Temporal to run the coordination: 32 badges per carrier, one Workflow per badge, and retries that ran on their own. Read the full story: https://lnkd.in/dWTTNrQs

Should we get into bowling?