Zero Dollar AI: Build Your First Agent in One Sitting
Details
## About
Most people think building an AI agent means paying for API access, juggling multiple provider accounts, and burning credits before they've even shipped anything.
**It doesn't have to.**
In this live session, we build a real, working AI agent from scratch — start to deployment — using only free-tier tools, in one sitting. No slides-only theory. No "we'll cover deployment next time." One build, start to finish, live in front of you.
---
## What You'll See Built, Live
- Connecting to **GitHub Models** with a single free key — no juggling API keys across providers
- **Function calling** to get structured, usable responses from an LLM
- A working **Streamlit** interface, wired up from scratch
- Real deployment to **Streamlit Cloud**
- Production-grade observability with **LangSmith** tracing
---
## Who This Is For
- Builders who want to move from "watching tutorials" to shipping something real
- Anyone who's been blocked from experimenting with AI because of cost or API friction
- Developers curious about agentic AI but unsure where to start
- Community members who want a repeatable build pattern they can take and extend on their own
You don't need prior agent-building experience. You do need curiosity and a GitHub account.
---
## What You Walk Away With
- A working understanding of how to access multiple LLMs through one free key
- A live-built agent you can fork, extend, and deploy yourself
- A repeatable pattern for function calling, structured output, and deployment
- Direct exposure to practices most tutorials skip: tracing, monitoring, real deployment
---
*Spots are limited to keep this hands-on and interactive. Bring your laptop — you'll be building, not just watching.*