Day 2: Production-Ready AI & Agentic Design

For Parallel inference, it still uses same/single cloud run instance for the agent??

We invite community to participate in this discussion. Thank you!

1 comment

Henry Ruiz

4 months, 3 weeks ago

Yes, it does. The function assumes the agent is already deployed (for example, on Cloud Run or somewhere), since it requires the agent’s public URL to run the inference.

If you’d like to review the implementation details, you can find it here:

https://github.com/vladkol/agent-evaluation-lab/blob/670f0a3386a0b3301146135eea4d8f9f618eae46/shared/evaluation/evaluate.py#L402

Priti yadav

4 months, 3 weeks ago