You’re offline. This is a read only version of the page.
Announcements
Hi everyone, 👋 From the AI Builder product group we want to hear from you.
👉Have you tried Text Generation in AI Builder already? How are you using it? For what use case? Is there anything you would like to see added or changed to make it even better?
👉We are interested in knowing about your AI Builder experience. Please share your story!
Thank you!
@Joel-Garrison — I just stumbled upon your post. I’m curious, what’s your end goal with the JSON file containing GPT-extracted text?
@Mehulge at Microsoft we keep always the security and privacy. So, the data is not used to re-train the models or to share with others. Here I paste a link where speak more about it.
Hi @plarrue
Happy to share with you two scenarios were I've work. These following scenarios are only using the Power Platform low code tools. The most important AI Builder to consume the ChatGPT model.
But as well we had some scenarios where we had to extend AI Capabilities using the Azure Open AI services and Azure Functions to consume, orchestrate and integrate the AI services to the Power Platform as a Custom Connector.
Sector: banking industry (just to check capabilities). The main idea here was to use it in combination of PVA and PA to have a better detection of the intent and entities contained in the user message. With that was enabled the "open questions" without having to do many configuration... having more intelligent chatbot. E.g : if users asks Q: I want to transfer 10 euros to [number account] the GPT will be able using a nice Prompt the intent in that case "transfer and will return a JSON object with the entities extracted to present an Adaptive Card with a form with the pre-filled data. Disclaimer: this is a workaround, this scenarios will be improved with a native NLU integration with PVA giving a more robust way.
Sector: insurance industry. Have a intelligent document process to manage judicial documentation. This documents are complex to read, with dynamic data and unstructured data. The IDP was in charge 2 main things, classify the document, extract the intent and extract the entities. Then this data was processes to insert/updated the information in a Dataverse. As well in case the GPT action is not able to extract some of the data then this document will be queued to a Canvas app to have a human validation. This extremely fast to implement thank to a accelerator, here is the link: Automate the processing of documents with the AI Builder prepackaged solution - Training | Microsoft Learn
Can you share with me privately your Environment ID or Org ID?
Is this a recurring issue, or it just happened once?
Ashish
Thank you Ashish and I will wait for your update.
Stay up to date on forum activity by subscribing. You can also customize your in-app and email Notification settings across all subscriptions.
Share
Report
All responses (
Answers (
7
