Temporal recently put out a wonderful demo of using Temporal to make prod-ready OpenAI agents, and I’ve seen many around asking a particular question about Temporal itself:

How do I stream responses from Temporal?

It’s a natural question, given how often SSE responses are used with LLMs. There are a few ways to approach this problem, but I will focus on a singular solution. I implemented my own Worker(s) -> SSE stream implementation about a year ago to handle fan-out notifications, and will share it with you here. I do not prescribe that this is the ideal solution, instead, this is meant to tease out some creative solutions from you, the reader.

In this post I’m going to share some tips on leading your AWS MediaConvert integration design with UX and SLA. I’ll do this by telling you a story about a fictional social media platform, VidBuddies, and its fictional CEO (my cat Patches).

Meet my cat Patches. She is an incredibly talented engineer and entrepreneur. Patches is building a social media video platform called VidBuddies where users (mainly other cats) can contribute content and interact with each other (think TikTok, or YouTube Shorts).

Have we talked about Temporal?

Look, your author is not one to shill for a product. In fact, most of the time I approach products with cantankerous skepticism. But Temporal is special. The “durable execution” paradigm it champions is something the industry sorely needs.

In this talk, I demonstrate how we handle large-scale video aesthetic analysis, with Temporal as the backbone orchestration mechanism. Wrangling a machine learning (ML) workflow across dozens of executors and endpoints.