Agentic AI System Development

1 Capstone Agent Design Refinement Refine your initial concept by specifying how your agent will reason, what memory it will require, and which tools it may need to complete its tasks effectively. 2 RAG and Retrieval Design Integration Decide whether your agent requires external knowledge or retrieval. You will define how retrieval, data grounding, and memory design will support more reliable outputs. 3 Tree-of-Thought Integration Plan Identify where structured reasoning can improve your agent. You will define how branching, evaluation, and selection could strengthen performance on more complex tasks. 4 Multi-Agent Architecture and Coordination Plan Extend your design into a multi-agent workflow. You will define specialized agent roles, communication patterns, and coordination logic to support more complex problem-solving. 5 Required Capstone Checkpoint 6.1: Safety Guardrails and Human Intervention Plan Specify how your system will be evaluated and monitored. You will define guardrails, logging needs, success criteria, and when human intervention should be used to improve safety and reliability. 6 Final Capstone Project Report and Presentation Complete, evaluate, and present your final agentic AI system via a report and video presentation. You will integrate the full architecture into a working solution and communicate your design decisions, results, and key takeaways. By the end of the program, you will have built an agentic AI system from the three options provided that demonstrates your ability to design, implement, evaluate, and communicate an end-to-end solution.

Project ID: 40591549