How luisnoronha/gen-ai-course works
This is a hands-on Generative AI course repository, not one deployed product. Its central Course 2 workflow is to start from an incident scenario, run an agent against a simulated exchange through tools, and inspect/evaluate the response in labs. Course 1 separately teaches LLM, retrieval, RAG, evaluation, and API workflows. The graph preserves these major branches; links are limited to calls visible in excerpts or references in the source index, and README-described capabilities are represented without assuming unshown wiring.
Main entry points: simulator.py Exchange Simulator · agent_runtime.py Agent Runner · tools.py Agent Tools · ui.py Shared Lab UI
Components 20
Course Material
- Fundamentals Sessions
sessions/course-01-ai-fundamentals - RAG Support
sessions/course-01-ai-fundamentals/rag_course_support.py - RAG Agent API
sessions/course-01-ai-fundamentals/07-agent-api/agent_service.py
Meridian Simulation
- Incident Scenarios
meridian/scenarios.py - Exchange Simulator
meridian/simulator.py - Exchange State
meridian/model.py - Metrics and Signals
meridian/signals.py - Simulation Rules
meridian/rules.py - Exchange Actions
meridian/actions.py - Fault Injection
meridian/faults.py
Agent Runtime
- Agent Tools
meridian/tools.py - Agent Runner
meridian/agent_runtime.py - Agent Memory
meridian/memory.py
External Interfaces
- MCP Servers
meridian/mcp_servers
Course 2 Labs
- Shared Lab UI
meridian/lab/ui.py - Security Exercises
sessions/course-02-agentic-ai/05-security/lab - Agent Evaluation
sessions/course-02-agentic-ai/06-evaluation/lab - Multi-Agent Lab
sessions/course-02-agentic-ai/07-multi-agent/lab - Incident Board
meridian/lab/board.py
Other components
- Learner
Connections 24
- Learner → Fundamentals Sessions: works through
README.md - Learner → Agent Runner: runs incidents
- Fundamentals Sessions → RAG Support: uses shared RAG
sessions/course-01-ai-fundamentals/04-rag-pipeline/starter/01_rag_baseline.py - RAG Agent API → RAG Support: reuses RAG
sessions/course-01-ai-fundamentals/07-agent-api/agent_service.py - Agent Runner → Incident Scenarios: loads
meridian/agent_runtime.py - Agent Runner → Exchange Simulator: initializes
meridian/agent_runtime.py - Agent Runner → Agent Tools: creates
meridian/agent_runtime.py - Agent Runner → Agent Memory: opens optional
meridian/agent_runtime.py - Exchange Simulator → Incident Scenarios: reads
meridian/simulator.py - Exchange Simulator → Exchange State: updates
meridian/simulator.py - Exchange Simulator → Metrics and Signals: generates
meridian/simulator.py - Exchange Simulator → Simulation Rules: applies
meridian/simulator.py - Exchange Simulator → Exchange Actions: applies
meridian/simulator.py - Exchange Simulator → Fault Injection: injects
meridian/simulator.py - Agent Tools → Exchange Simulator: operates
meridian/tools.py - Agent Tools → Exchange Actions: invokes
meridian/tools.py - Metrics and Signals → Exchange State: derives from
meridian/signals.py - Security Exercises → Agent Runner: runs agents
sessions/course-02-agentic-ai/05-security/lab/agent.py - Agent Evaluation → Agent Runner: evaluates agents
sessions/course-02-agentic-ai/06-evaluation/lab/agents.py - Multi-Agent Lab → Incident Board: updates display
sessions/course-02-agentic-ai/07-multi-agent/lab/app.py - Multi-Agent Lab → Shared Lab UI: uses shared UI
sessions/course-02-agentic-ai/07-multi-agent/lab/app.py - Security Exercises → Shared Lab UI: uses shared UI
sessions/course-02-agentic-ai/05-security/lab/app.py - Agent Evaluation → Shared Lab UI: uses shared UI
sessions/course-02-agentic-ai/06-evaluation/lab/app.py - Agent Runner → MCP Servers: supports optional