---
title: "Comment by Dr Srinivas Padmanabhuni on \"How to Build Reliable AI Function Calling for Business Applications\" | Hashnode"
description: "Good point that schema validation checks structure but not whether the product exists or the user is allowed to see it; that gap is easy to miss. A check I'd ad"
url: "https://hashnode.com/posts/how-to-build-reliable-ai-function-calling-for-business-applications/6afefb8af9d7c9d4ae00c8a2/comment/6aff094229c4bc505b3f915e"
---

[Discussion](https://hashnode.com/posts/how-to-build-reliable-ai-function-calling-for-business-applications/6afefb8af9d7c9d4ae00c8a2)

Comment [How to Build Reliable AI Function Calling for Business Applications](https://hashnode.com/posts/how-to-build-reliable-ai-function-calling-for-business-applications/6afefb8af9d7c9d4ae00c8a2)

[Dr Srinivas Padmanabhuni](https://hashnode.com/@srinivaspadmanabhuni)

[1d ago](https://hashnode.com/posts/how-to-build-reliable-ai-function-calling-for-business-applications/6afefb8af9d7c9d4ae00c8a2/comment/6aff094229c4bc505b3f915e)

Good point that schema validation checks structure but not whether the product exists or the user is allowed to see it; that gap is easy to miss. A check I'd add to the workflow test is reading the saved trace afterwards: does the assistant's final answer match what the tool really returned, for example saying a quotation was prepared when prepare\_quotation errored? AiEnsured agentest runs that kind of trace check (pip install agentest-ai, 165 tests), and developers are welcome to enhance and contribute.
