CQUniversity Unit Profile.

• You are required to write an individual report (1,500–2,000 words) that includes: 

1. Research Component (approx. 600–700 words): Investigate current GenAI
2. Self‑Evaluation Component (approx. 500–600 words): Reflect on your own use of GenAI tools during Weeks 1–4 of the course. This section must be personal and evidence‑based.
3. Critical Discussion Component (approx. 250–300 words)
4. Conclusion (approx. 100–150 words) 

Research Quality 30% 
Depth, accuracy, and critical evaluation of GenAI tools across the SDLC 
Self‑Evaluation 30% 
Depth of reflection, personal insight, and use of evidence 
Critical Discussion 20% 
Quality of argument, ethical awareness, and personal perspective 
Structure & Writing Quality 10% 
Clarity, coherence, academic style, and referencing 
Academic Integrity & Originality 10% 
Use of personal evidence, proper citation, and authenticity 

In Assessment 2, you will design, implement, integrate, and evaluate a single intelligent agent AI Application. The details of the application's functional specification are left to your imagination. 
You are required to: 

• Build a LangChain‑based agent
• Implement internal state management
• Integrate the agent into FastAPI endpoints
• Write a comprehensive pytest suite
• Evaluate agent behaviour and reliability
• Produce a technical report documenting design decisions

This assessment emphasises the software engineering discipline, not just AI prompting.

Agent Implementation 25% 
Correct use of LangChain, tools, prompts, and execution flow 
State Management 15% 
Correct design, updates, logging, and integration 
Endpoint Integration 15% 
FastAPI correctness, validation, and error handling 
Testing Suite 20% 
pytest coverage, mocking, integration tests 
Evaluation & Report 20% 
Depth of analysis, clarity, originality 
Professionalism & Integrity 5% 
Code quality, code documentation, and AI declaration 

In Assessment 3, you will design, implement, and evaluate a basic dule‑agent system using the concepts introduced in Weeks 6-9: 

• Agent communication
• Message passing
• Orchestration
• Coordination
• Resilience and guardrails 

You will extend your  AI application in assessment 1 into a project by adding two specialised agents that collaborate to complete a task. 
Examples include: 

• Planner Agent + Worker Agent
• Summariser Agent + Reviewer Agent
• Classifier Agent + Validator Agent

 
This assessment emphasises communication design, coordination logic, and system‑level behaviour, not just individual agent performance. 

1. Agent Design & Implementation Roles, behaviour, structure 20%
2. Communication Protocol Message structure, validation, logging 20%
3. Orchestration & Coordination Routing, control flow, iteration control 20%
4. Resilience & Guardrails Timeouts, validation, fallback logic 15%
5. Evaluation & Technical Report Analysis, clarity, originality 20%

6. Professionalism & Integrity<br>Code quality, documentation, AI declaration 5%

In this Assessment, you will design, implement, evaluate, and deploy a full multi‑agent Application that integrates:

• Core task management functionality 
• Robust software design and architecture 
• A multi‑agent workflow with at least 3 agents + orchestrator 
• Deployment, observability, and resilience 

This is your capstone: it should demonstrate end‑to‑end engineering, not just prompting.

This assessment evaluates your ability to: 

• Translate requirements into a coherent system architecture, 10%
• Design specialised agents with clear roles and interfaces, 10%
• Implement multi‑agent workflows with structured communication, 20%
• Apply resilience, guardrails, and observability, 20%
• Deploy a working system in a realistic environment, 20%
• Critically evaluate system behaviour, including failure cases, 15%
• Produce professional documentation and a demonstration, 5%