EC402: Applied AI & Autonomous Agents

Build AI systems with intention, not cargo cults

Pillars

• Agent loops (plan→act→observe→reflect)
• Tool use, retrieval, memory strategies
• Evaluation, guardrails, human-in-the-loop design
• Orchestrating multi-agent swarms for hackathons
• Deploying agents into production systems

Projects

1. Research copilot with long-term memory
2. Multi-agent hackathon assistant
3. Production-ready workflow automation

Outcomes

• Architect agents end-to-end
• Decide when to use RAG vs fine-tuning vs tool calling
• Explain risk, alignment, and observability tradeoffs

Difficulty: ⭐⭐⭐⭐⭐ | Time: 18 hours | Prereqs: EC401 + EC302 + EC301