Employee Report

To access secure data in private network inside azure datalake using managed endpoint and Ingest, transform and present it in PowerBI. Finally expose the data to AI for novel output of user queries.

Comment 1

Inspiration

The recent shifts in the Data & AI landscape inspired me to join this hackathon. I wanted to experiment with cutting-edge technology, specifically Microsoft Fabric for data and Langchain for AI. So, I set myself a challenge to work with both.

What it does

I developed an employee management tool that enables organizations to identify the number of domain experts within specific cities with their own data. The tool accepts voice input from users, which is then converted into text and subsequently transformed into an SQL query with the help of AI. The generated SQL query is executed on Microsoft Fabric SQL Lakehouse and give back the result to user.

Challenges we ran into

Initially, I was uncertain about the feasibility of connecting Microsoft Fabric with AI agents. Through experimentation with SQL endpoints, I discovered an effective approach to link a Python LangChain application. This enabled me to leverage SQL agents for querying Fabric successfully.

The project's data is securely stored in a data lake behind a VNET, which posed initial challenges for establishing a secure connection. Configuring and optimizing this setup took some time but was essential for ensuring robust security and seamless data access.

Accomplishments that we're proud of

I had been feeling that I hadn’t worked in AI for a while. Participating in this hackathon made me really happy, as I got to work with AI and explore the cutting-edge tool, Microsoft Fabric.

What we learned

In Microsoft Fabric: Fabric Workspaces, Fabric Experiences Onelake SQL endpoint Medallion architecture Creating Facts and Dimension tables Fabric PowerBI Connecting delta lake to API based application Using Machine learning inside Fabric (Connected hackathon model inside Fabric Notebook to get results on top of organisation data)

AI: Langchain, Langchain Agent, Connecting to models cognitiveservices cognitiveservices speech

What's next for Employee Report

The primary objective of this project is to implement it within our organization, enabling employees to switch roles every six months or less. This approach aims to enhance their learning experience by allowing them to gain hands-on exposure to end-to-end projects. In the future, I plan to expand the tool’s capabilities, making it a key resource for continuous skill development and seamless role transitions across the organization.

Built With

Share this project:

Updates