AI Evenings LDN: Building Agentic Systems with Google

As AI agents move into production, teams need systems that can reason, act, and be trusted. In this session, Alvaro Silva-Santisteban shares how agentic systems are designed in practice, from architecture to governance and deployment.

• Anatomy of an Agent - A practical breakdown of what actually makes an AI agent work. Alvaro deconstructs the core stack, the Brain (models), Memory (sessions and long-term context), and Hands (function calling and tools), and how these pieces come together to enable reasoning and action.
• Choosing the Right Agent Architecture - Not every problem needs the same kind of agent. Learn how to distinguish between retrieval-based agents and action-oriented agents, and how to align each approach to real business needs without over-engineering solutions.
• Governance, Trust, and Agentic Workflows - Agentic systems introduce new risks alongside new capabilities. This section covers mitigating hallucinations, implementing human-in-the-loop safeguards, and defining the role of business users, plus a walkthrough of agentic workflows using tools like Deep Research agents, NotebookLM, and AI Studio.

Know what’s Happening Next — before everyone else does.

Get App

Host or Publisher BrainStation