I'm Shiva, a Group Product Manager at Uber, where I lead the Communications Platform & Intelligence team. It's the system behind hundreds of billions of real-time interactions a year — voice, chat, push, email, SMS, and WhatsApp — across Rides, Eats, Driver, and Merchant, used by hundreds of 

teams. Today I manage and optimize a $100M+ global platform, balancing efficiency, reliability, and user experience at scale.

My work sits where AI meets distributed systems and real human conversations: turning LLMs, speech-to-text, and real-time inference into production-grade platforms. Lately that means LLM-driven intent detection and real-time interventions across chat and voice — figuring out what a conversation is really about and what should happen next, from a support answer to stopping fraud in the moment.

That's also why I'm teaching this. I spend my days in the gap between AI hype and what actually ships, earns trust, and survives security, legal, and compliance review at global scale. Most "autonomous agent" demos never cross that gap — this lesson is the honest version of how you do.

Before the comms platform, I led Adtech and Martech products at Uber, and was Head of Product at Ampush building multi-channel marketing and data systems. I've also been an instructor and featured speaker at Product School for 8 years, helping the next generation of PMs build with AI honestly.