Running an Aesthetic Analysis ML System on Temporal
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.