AWS Tech Talk

Date

Wednesday, October 14, 2026

Time

6PM - 7PM

Location

Fontana Labs 2000Get directions →

Come join us to hear Siddharth Garava, Software Engineer at AWS talk about how to build AI Agents! Calling an LLM is easy. Building an agent that can choose tools, remember progress, recover from errors, and prove that it completed a task requires real software engineering. In this live-coding session, we will start with a basic model call and progressively build a tool-using investigation agent. We will add structured tool definitions, a planning loop, state management, error handling, and a small evaluation suite that tests whether the agent selected the correct tools and gathered enough evidence. Along the way, we will examine where the model should make decisions and where deterministic code should remain in control. Students will leave with a working reference implementation and practical patterns they can reuse in hackathons, personal projects, and production AI applications.