Performance Issue with Agents - Synchronous API Call

[puneetsharma7445]

Please read this first

• Have you read the docs?Agents SDK docs - Yes
• Have you searched for related issues? Others may have had similar requesrs - Yes but haven't found similar one

Question

I'm working on a use case where a user uploads an Excel file containing around 20 product details. The backend API processes this request by invoking an agent, which reads the file content and returns the product details in a structured JSON format. However, we're facing performance issues— the agent takes 20-33 seconds to process the data, leading to gateway timeout errors.

Am I approaching this incorrectly? Can agents be effectively used in a synchronous REST API call? Any suggestions on improving performance or handling this better would be greatly appreciated!

[rm-openai]

That sounds possible, though slower that I would expect if it's a small excel file. Can you try it in the openai playground and see how it does? https://platform.openai.com/playground/prompts

Options to make it faster:

1. Use async operations. i.e. return a 200 immediately, and then push an update to the client when done.
2. Use server-sent events. This is what most apps (e.g. chatgpt) do, keeping a one way connection open and pushing events to the client.
3. Use a longer server timeout.

[puneetsharma7445]

Thanks for your response @rm-openai .
I will try it on playground and share the findings.

[github-actions]

This issue is stale because it has been open for 7 days with no activity.

[github-actions]

This issue was closed because it has been inactive for 3 days since being marked as stale.