Summary
- Covers using GITHUB COPILOT to fix bugs in the eyes of a new developer
- How to build a LLM model using your own dataset
In order to make applications intelligent/AI capable - they need to be modern i.e. micro services, API’s, CI/CD etc.
Need clean, pure data to drive decisions - no mistakes
Example…
Five steps…
Q: How much research has been done to determine that end users want to interact with a system using natural language vs a point and click interface?
A: I recommend to follow Jaime Teevan- Chief Scientist & Technical Fellow at Microsoft for the dozens of interesting researches on future of work & HCI
https://www.linkedin.com/in/jaimeteevan/
Every developer needs to be 10X more vigilant on the automated code output - it contains bugs!
Try: VSCODE Insider Preview - Agent Code preview
Azure AI Foundry - Chat Playground
GPT 3.5 TURBO model & added custom data
Can use it to filter out questions
We encourage everyone to join our next 2 sessions on Security
Data Security for AI
https://msevents.microsoft.com/event?id=12142814910245&ocid=lp_pg446902_gdc_comm_az
Code Security Reinvented: Navigating the era of AI
https://msevents.microsoft.com/event?id=1314627257308&ocid=lp_pg446902_gdc_comm_az
