Module 11: Enterprise AI Project

Goal

Combine RAG, agents, evaluation, cloud architecture, security, observability, and product thinking into one production-style system.

Required Capabilities

• Multi-user authentication.
• Role-based authorization.
• Document ingestion.
• Retrieval with citations.
• Tool calling.
• Read and write workflows.
• Human confirmation for risky actions.
• Audit logs.
• Evaluation pipeline.
• Monitoring and cost tracking.
• Containerized deployment.
• Cloud architecture plan.
• Architecture decision records.

Study Tasks

• Select a realistic domain process.
• Write requirements and acceptance criteria.
• Create system architecture diagrams.
• Implement core backend services.
• Implement RAG.
• Implement agent tools.
• Add evaluation.
• Add security controls.
• Add deployment documentation.
• Record architecture decisions.

Project

Build projects/09-final-enterprise-ai-system/.

Tips

• Do not build everything at once. Ship vertical slices.
• Keep the demo path reliable.
• Document trade-offs as you make them.
• Reduce scope before reducing quality.

Completion Checklist

• The system has a working end-to-end demo.
• The system has documented security controls.
• The system has measurable evaluation results.
• The system has deployment instructions.
• The system has architecture diagrams and ADRs.