Build Your Own ChatGPT in Azure
BeConnected day 12 + CollabDays 2024
30/05/2024 UCI Conference Centre Assago - Milan
https://www.beconnectedday.it/bcd12-a...
Build Your Own ChatGPT in Azure
Michael Noel: Partner at Convergent Computing
Large Language Models (LLMs) like OpenAI’s ChatGPT have emerged as a disruptive force for organizations all over the world. The ability to search through disparate content using natural language queries and to be able to intelligently pull together useful information that would ordinarily be siloed is incredibly useful to modern organizations. Those same organizations are naturally reluctant to upload their own proprietary information into public LLMs and are often limited to using these tools against public datasets.
As part of Microsoft’s partnership with OpenAI, several new offerings have been introduced to Microsoft’s Azure cloud services that allow organizations to use a ChatGPT engine to query their own proprietary content and to create valuable chat engines that employees can use to quickly and efficiently query and reference internal knowledge. These tools are powerful and easy to configure and can be used immediately by organizations of any size, not requiring any special licensing above and beyond minimal costs for the running Azure infrastructure.
This session demonstrates the step-by-step process of deploying a ChatGPT chat engine in Azure, focusing on all steps needed to build the environment from scratch, including loading content, creating the search indexes, and provisioning the web application to run the engine on.
• View in detail the steps necessary to build your own ChatGPT environment that queries off your own data in your own isolated Azure instance
• Understand how to maintain data within the environment and how to upload and update the indexes and documents in the platform
• Understand the security ramifications of using LLMs and how to properly isolate your environment
Deck: https://api.runevents.net/api/assets/...