Overview

Designed and delivered a hands-on workshop that teaches engineering teams how to build production-ready AI agents from scratch, using a progressive commit-based learning approach.

Problem

Most AI agent tutorials are either too theoretical or skip the fundamentals. Engineers needed a structured, hands-on path from zero to a working AI agent.

Solution

Built the workshop as a series of 13 progressive git commits, each introducing a new concept — from basic API calls to tool use, memory, and multi-step reasoning. Included a full speaker script for reproducibility.

Tech Stack

• TypeScript — Workshop codebase
• Claude API — Anthropic's language model
• OpenAI (gpt-4o) — Alternative model support

My Role

• Designed the curriculum and progressive learning path
• Built all 13 workshop stages as git commits
• Wrote the full speaker script and facilitation guide
• Delivered the workshop to engineering teams

Impact

Enabled engineers to build production-ready AI agents from scratch, with a reusable workshop format that can be delivered by any facilitator.