This has been very good. I have been able to setup a file like yours with sample queries and take that all the way through to building an index and using it in the playground. I need to do a lot more work on the sample queries, and incorporate details from other tables in the database. There is a "Deploy" button in AI Foundry. What is this really intended for? Where does the "code" live. In addition if I deploy this way, how do I go from asking a question, and the response being SQL to the response being a commentary on the data returned? In your demo, you returned revenue over a period of time and charted the result. Can you take this a step further and get an AI commentary on the response to talk about trends in the data? for example... please compare revenue for December 2023 and December 2024 across the top 5 products and comment on trends taking into account seasonal variance in the product sales.
The .txt file of SQL queries you have... how do you process a update to that file... lets say I have added more questions and SQL to the file... what is the next part of the process I need to follow to get a updated Index?
hi, still I am not able to find a way for deploying the model via api. For example if I want to deplot gpt 4o via code instead of using azure ai studio?
How can i connect cutomers to the chat ? I mean for personalization user 1 should get different results as per his taste and preference, whereas user 2 get different results as per his defined taste and preference
One way to achieve this is by passing table metadata (via Azure Search or directly in the prompt) and updating the prompt dynamically based on the user's context in your backend app. Additionally, building a validation layer ensures the generated SQL queries are accurate and restricted to each user's data.
Nice sharing
Great keep going waiting for new one
Thanks for sharing Sir 🙏
Thanks for sharing 🙏
Good
Good , keep it up!
Nice knowledge 🎉🎉
This has been very good. I have been able to setup a file like yours with sample queries and take that all the way through to building an index and using it in the playground.
I need to do a lot more work on the sample queries, and incorporate details from other tables in the database.
There is a "Deploy" button in AI Foundry. What is this really intended for? Where does the "code" live.
In addition if I deploy this way, how do I go from asking a question, and the response being SQL to the response being a commentary on the data returned?
In your demo, you returned revenue over a period of time and charted the result. Can you take this a step further and get an AI commentary on the response to talk about trends in the data? for example... please compare revenue for December 2023 and December 2024 across the top 5 products and comment on trends taking into account seasonal variance in the product sales.
The .txt file of SQL queries you have... how do you process a update to that file... lets say I have added more questions and SQL to the file... what is the next part of the process I need to follow to get a updated Index?
hi, still I am not able to find a way for deploying the model via api. For example if I want to deplot gpt 4o via code instead of using azure ai studio?
How can i connect cutomers to the chat ? I mean for personalization user 1 should get different results as per his taste and preference, whereas user 2 get different results as per his defined taste and preference
One way to achieve this is by passing table metadata (via Azure Search or directly in the prompt) and updating the prompt dynamically based on the user's context in your backend app.
Additionally, building a validation layer ensures the generated SQL queries are accurate and restricted to each user's data.