Madhuri Somara, Senior Product Manager at Microsoft, joins Federico to unpack how she builds AI agents that actually help people, not just impress on paper. Fresh off being honored with the 2025 Product Leader Award by Products That Count, Madhuri traces her path from coding and business analysis to product leadership, and why empathy, rigorous evaluations, and clear user value are her north stars.

She describes the vision and real-world impact of Microsoft’s Autonomous Agents (announced at Ignite 2024), zooming in on the Case Management Agent that automates the entire lifecycle from case creation to closure while preserving human supervision. For Madhuri, customer shadowing, sentiment reading, and removing “small frictions” (like a few extra clicks) compound into big wins that frontline teams feel every day.

A throughline in the conversation is AI Evaluations (AI Evals). Madhuri explains why intuition and basic testing aren’t enough when models act on behalf of enterprises and customers. Strong evals and golden datasets build confidence, keep behavior within guardrails, and ensure products behave as intended over time — akin to behavior-driven development but for AI behavior. She also clarifies the nuance between “human-in-the-loop” (blocking dependency) and human supervision (oversight with autonomous progress), and how trust, reliability, and safety guide the right choice.

Beyond shipping features, Madhuri emphasizes UX as behavior design — predictable, accessible, and consistent interfaces that reflect how people actually work. She shares how collaboration between product, design, and engineering yields clearer requirements, fewer back-and-forths, and more predictable delivery.

On responsible AI, she’s pragmatic: use AI where it clearly adds value; don’t force it. Balance innovation with adoption and real user pain points. Looking ahead, she predicts agentic AI will reshape work across customer service, sales, marketing, IT, and more — freeing humans for higher-judgment, creative tasks. Tools like Copilot already remove mental load (drafts, comparisons, bookings), but she stays cautiously optimistic about privacy, reliability, and security.

Madhuri also spotlights community and growth: she mentors university students, champions diverse voices through Women in AI Ethics™, and advises aspiring AI PMs to network intentionally and stay curious — be “learn-it-alls,” not “know-it-alls.” Finally, for women and underrepresented talent, she offers practical encouragement: build relationships, ask questions across disciplines, and keep tying ideas back to real customer value.

About Madhuri Somara:

- https://www.linkedin.com/in/madhurisomara/

About Federico Ramallo ✨👨‍💻🌎

🚀 Software Engineering Manager | 🛠 Founder of DensityLabs.io & PreVetted.ai | 🤝 Connecting 🇺🇸 U.S. teams with top nearshore 🌎 LATAM engineers

- 💼 https://www.linkedin.com/in/framallo/

- 🌐 https://densitylabs.io

- ✅ https://prevetted.ai

🎙 PreVetted Podcast 🎧📡

- 🎯 https://prevetted.ai/podcast

- 🐦 https://x.com/PrevettedPod

- 🔗 https://www.linkedin.com/company/prevetted-podcast

00:00 Introduction to Madhuri Somara and Her Journey

04:30 Pivotal Moments in Career Development

08:40 Experiences at Avanade and GS1 US

13:00 Launching Autonomous Agents at Ignite 2024

18:15 Challenges in AI Product Management

23:01 The Importance of Customer Feedback

26:54 Finding Balance in Product Management

28:46 The Importance of AI Evaluations

31:46 Navigating AI Capabilities and Responsible Use

35:49 Advice for Women in AI Product Leadership

39:08 Future Impact of AI and Automation

40:58 Skills for Future Product Managers

45:49 Cautious Optimism for AI's Future