Michigan .NET Users Group (MI .NET)

Agenda

• 5:30 - 6:30pm Networking
• 6:30 - 7:30pm Presentation
• 7:30 - 7:45pm Closing

Abstract
Retrieval-Augmented Generation (RAG) has quickly become one of the most practical approaches for building AI applications, yet many developers treat it as a “black box.” In this talk, I peel back the layers and dive into the math and mechanics that make RAG work.

I begin with a brief overview of RAG and then focus on its two core components: embeddings stored in a vector database and the retrieval of relevant data using cosine similarity. I explain how text is transformed into embeddings, which are arrays of numbers that capture semantic meaning, and compare different methods of generating embeddings, highlighting why vector representations are essential in this context.

Next, I explore cosine similarity as the retrieval mechanism. By plotting embeddings in a 3D graph, I show how measuring the angle (cos θ) between vectors determines relevance.

Finally, I walk through a live coding demo to build a simple RAG pipeline, showing how embeddings are stored in a vector database and retrieved in real time.

Bio
Dev is a detail-oriented Software Engineer and GenAI Evangelist passionate about building scalable, cloud-based applications and AI-powered solutions. With a strong focus on engineering best practices and client collaboration, he delivers tailored solutions that meet unique needs. He channels that same passion into helping developers adopt GenAI technologies through technical blogs, videos, and conference talks.

Agenda

• 5:30 - 6:30pm Networking
• 6:30 - 7:30pm Presentation
• 7:30 - 7:45pm Closing

Abstract
AI agents are everywhere - but how do they actually work in a real data platform?

In this session, we’ll go beyond chatbots and build a practical, enterprise-ready multi-agent system using Databricks Agent Bricks. You’ll see how structured data (SQL tables, Delta Lake, Genie Space) and unstructured documents (PDFs, invoices, call logs, reviews) can be orchestrated through a supervisor agent to answer complex business questions.

Using a realistic enterprise demo, we’ll walk through:

• Extracting invoice data from PDFs and images into structured tables
• Grounding answers in product manuals and compliance documents
• Analyzing customer sentiment from call transcripts and reviews
• Routing tasks across specialized agents using a supervisor
• Connecting everything back to SQL analytics and dashboards

This is not theory — this is a working architecture that connects your existing SQL data estate to modern AI workflows.

If you work with SQL, data engineering, or analytics and want to understand how agentic AI fits into the real enterprise stack, this session is for you.

Bio
Mou Rakshit is a Senior Manager of Data, AI & Analytics Engineering at Accenture/Avanade and a Principal Data & AI Solution Architect specializing in building modern, unified, AI-ready data platforms using Microsoft Fabric and Databricks. She leads enterprise-scale data strategy and platform architecture initiatives that enable organizations to move from fragmented analytics environments to governed, high-performance lakehouse ecosystems.

Her work focuses on designing scalable medallion architectures, implementing robust governance models, enabling real-time intelligence, and operationalizing AI and advanced analytics across the enterprise. Mou bridges executive strategy with hands-on technical execution, ensuring that modern data platforms are production-ready, resilient, and aligned to business outcomes.

She holds a Master of Science in Computer Science from Wayne State University and a Master’s in Health Services Administration from the University of Michigan. A Databricks Champion and frequent industry speaker, Mou shares practical frameworks and real-world lessons on architecting unified data foundations that power AI at scale.

Agenda

• 5:30 - 6:30pm Networking
• 6:30 - 7:30pm Presentation
• 7:30 - 7:45pm Closing

Abstract
Many healthcare organizations live in a world of hybrid systems: half modern APIs, half legacy apps. Instead of replacing everything, what if you could connect and structure the data you already have?

This session explores how to unify siloed systems through smart metadata, semantic layers, and dependency-driven queries. You’ll learn strategies to evolve existing systems gradually—without downtime or total rewrites.

Bio
John Sly is a software developer and data engineer specializing in large-scale ETL systems and healthcare data integration. With over a decade of experience in PHP, PostgreSQL, and enterprise data workflows, he bridges legacy architectures with modern, automated solutions that improve reliability and transparency. His talks focus on practical engineering—what really works when moving complex data at scale and delivering something impactful at the end.

Agenda

• 5:30 - 7:30pm Holiday Party

Details
Join this joint event for a festive Holiday Party! Connect, unwind, and celebrate the season with fellow members. Food, drinks are available for purchase directly from the restaurant. Feel free to bring your favorite board games.