Finance Lab Session 10-Power BI, Fabric planning and AI Agents for Finance

Description
1. Background
This resource captures the key insights from Session 10 of the Future of Finance Lab, featuring data consulting CEO Khaled Chowdhury and hosted by Lance Rubin. The session explores why AI will create more Excel and Power BI usage — not less — and how Microsoft Fabric's semantic model layer provides the ideal foundation for production-ready AI agents in finance.

Included are practical frameworks for understanding the shift from report-building to agent ownership, a breakdown of how Power BI's built-in metadata advantage makes it superior to generic RAG deployments, and a cost-effective entry path to Fabric (pay-as-you-go, F2 capacity). The resource also covers Claude vs Copilot use cases and why domain expertise matters more than ever for finance professionals working with AI.


Suitable for finance managers, FP&A professionals, CFOs, and accountants exploring AI-augmented workflows.


3. What's Included:

• Session summary document (key insights, frameworks, and practical takeaways)
• Tool comparison matrix: Claude vs Copilot vs Gemini vs ChatGPT — finance use cases
• Microsoft Fabric entry path guide: pricing tiers, pay-as-you-go setup, F2 Copilot access
• Checklist: "From Report Builder to Agent Owner" — 10 steps for finance teams
• List of tools and resources mentioned in the session with links

4. Who Should Use: 


Finance professionals, FP&A analysts, management accountants, and CFOs who want to understand how AI agents, Microsoft Fabric, and Power BI semantic models are changing the finance function — and how to start experimenting without a large upfront investment.


5. Visuals/Imagery Guidance:

• Screenshot of the Power BI "prep data for AI" feature in Desktop
• Diagram showing the three-layer architecture: Data Source → Power BI Semantic Model → Data Agent → End User
• Side-by-side comparison of the star schema outputs from Claude vs Copilot (as discussed in session)
• Session slide showing the AI-Powered Accountant module structure