Azure AI Document Intelligence : Latency Issue

Sidharth Chauhan 25

I am using the Node SDK of @azure/ai-form-recognizer in my project and am experiencing latency issues. While debugging with New Relic, I found transaction traces related to formrecognizer/documentModels. Within these traces, there are internal calls made by the Node SDK, and I identified two calls to timers.setTimeout, each taking approximately 5 seconds.

I would like to reduce this timeout duration to improve the overall latency of our application. Below are the request IDs for your reference:

35a361e2-eb60-4a79-b1a2-8e25b7613b4f
3f835618-24ec-492b-9eb9-bca2bf0279c9

Could you please assist in understanding the source of these setTimeout calls and provide guidance on how to reduce this time?
Screenshot 2024-07-09 at 12.24.25 PM

santoshkc 9,565 Reputation points Microsoft Vendor

2024-07-09T08:56:40.16+00:00
Hi @Sidharth Chauhan,

Thank you for reaching out to Microsoft Q&A forum!

Document Intelligence is a multitenant service where latency is the amount of time it takes for an API server to handle and process an incoming request and deliver the outgoing response to the client. Occasional variability in latency and performance is inherent in any microservice-based, stateless, asynchronous service that processes images and large documents at scale.

Although we're continuously scaling up the hardware and capacity and scaling capabilities, you might still have latency issues at runtime.

Did you got any chance to check this: Best practices to lower the latency

Here are some guidelines to follow to reduce the latency:

Use a smaller model: Smaller models have fewer parameters, which means they can be loaded faster and run faster.

Use a pre-built model: Try using a pre-built model. Pre-built models are optimized for performance and can be faster than custom models.

Use a GPU: Try using GPUs, which can accelerate the processing of models and reduce the latency of document recognition.

Use batch processing: If you are processing multiple documents, try using batch processing. Batch processing can reduce the overall processing time by processing multiple documents at once.

I hope you understand! Thank you.
Sidharth Chauhan 25 Reputation points

2024-07-09T11:59:58.44+00:00

I already had consider all the best practices which you mentioned here.
Link of"Best practices to lower the latency" has different page - Lower speech synthesis latency using Speech SDK.
Can you please debug it with requestIds i had shared. It is not related to network latency , it is actually taking this much time on azure side only.
santoshkc 9,565 Reputation points Microsoft Vendor

2024-07-10T06:55:21.2066667+00:00
Hi @Sidharth Chauhan,

Sorry for the trouble you are facing. I recommend reporting this issue to the Azure Service support team. They will be able to investigate the issue further and provide a more targeted solution. You can report the issue by following these steps:

Go to the Azure portal and navigate to your Document Intelligence Services resource.

Click on the "Support + troubleshooting" tab.

Fill out the required information, including a detailed description of the issue and any steps you have taken to troubleshoot it.

Submit the support request.

The Azure support team will review your request and provide assistance as soon as possible.

Thank you.

Share via

Azure AI Document Intelligence : Latency Issue

Your answer