Speakers

Mariano Cocirio

Product Manager, DeepMind

Product Manager at Google DeepMind, focusing on agentic capabilities within Gemini.

Building Agents with Gemini

An inside look at building production-grade agents with Gemini — architecture decisions, agentic capabilities, and the real trade-offs of shipping AI at Google DeepMind.

Sasha Denisov

CTO, Brainform.ai

CTO at Brainform.ai with 20+ years of experience architecting scalable enterprise systems. Google Developer Expert for Cloud, AI, Firebase, Flutter, and Dart. International speaker at 30+ conferences worldwide.

Building Voice AI Agent That Listens, Understands, and (Most Importantly) Sells

Voice commerce is moving beyond chatbots. Google's Multimodal Live API enables 600ms-latency real-time conversations with native audio, multimodal understanding, and natural interruptions via WebSocket streaming. We built an AI shopping assistant that combines Live API with RAG — accessing product catalogs and knowledge bases in real-time through function calling to guide customers through purchases by voice. This talk covers our journey from prototype to production: WebSocket architecture, RAG pipeline design, function calling patterns, and handling edge cases.

Yusef Mohamadi

Software Engineer, Delivery Hero

Software engineer with a Master's in AI, combining both to build intelligent operational systems. Currently at Delivery Hero, driving speed and efficiency with AI orchestration at scale.

Agentic Operations: Driving Speed and Efficiency with AI Orchestration

Traditional automation is no longer enough to keep pace. The future belongs to agentic operations — and we have proved it works. In this session, we share how Delivery Hero successfully transitioned from static workflows to dynamic, AI-driven orchestration, achieving extraordinary operational results. Discover how autonomous AI agents can seamlessly coordinate tasks, make real-time decisions, and eliminate bottlenecks at scale.

Anthony Alaribe

Founder, Monoscope

Founder of Monoscope, building tools that make software operations more autonomous and reliable.

Runbooks That Think: Executable Checklists with AI in the Loop

Operational runbooks are static documents that demand constant human attention. What if they could reason, adapt, and execute autonomously? This talk explores how to turn passive checklists into dynamic, AI-driven workflows — covering the architecture, the failure modes, and the practical patterns for keeping humans in the loop when it actually matters.