EC302: Databases, Knowledge & Memory Systems

Designing Reality into Tables, Docs & Graphs

Focus Areas

• Relational modeling, normalization, joins
• Document & key-value stores, when schemaless wins
• Indexing, transactions, consistency models
• Knowledge graphs, vector stores, hybrid systems
• Retrieval design for AI agents & hackathons

Projects

1. Model a messy real-world domain multiple ways
2. Design retrieval pipelines for agent memory
3. Optimize queries and indexes under constraints

Outcomes

• Treat databases like engineered long-term memory
• Confidently choose the right persistence strategy
• Build data layers that feed AI workflows

Difficulty: ⭐⭐⭐⭐☆ | Time: 15 hours | Prereqs: EC102 + EC301