➡️ Next - Capstone Project Submission Guidelines
This is the core projects in the Mastering AI Agents Certification Program. These projects are the main deliverables you'll complete to earn your certificate of completion.
Each project is designed to apply what you've learned in that program through hands-on implementation. This project wraps up Mastering AI Agents program by having you build a working multi-agent system that demonstrates the key concepts from Unit 1-4: transitioning from workflows to agents and designing multi-agent collaboration using orchestration frameworks.
In this project, you'll design and implement a multi-agent system that demonstrates your mastery of the core Mastering AI Agents concepts:
You'll build a multi-agent system that incorporates the key learning objectives from Mastering AI Agents program. Your system must include:
1. Multi-Agent System (minimum 3 agents)
2. Tool Integration
Get a quick overview of what’s expected in the Mastering AI Agents project, along with a sample multi-agent system to inspire your own implementation.
Your multi-agent system should solve a meaningful problem that benefits from multiple agents working together. The problem should be complex enough to require coordination, but scoped to be doable in a couple of weeks. Here are some example directions to consider:
Build a system that helps users improve how their AI/ML project is presented for public sharing. The user provides a GitHub repo and (optionally) a brief project description. Your system returns practical suggestions to enhance discoverability, clarity, and completeness.
Sample tasks your system might perform:
🧠 Agents you might consider (not all are needed):
Repo Analyzer – parses and interprets README, code, and structure
Metadata Recommender – suggests tags, categories, and keywords
Content Improver – proposes better titles, summaries, or intros
Reviewer / Critic – checks for missing sections or unclear areas
Fact-Checker – verifies claims or suggestions against actual repo content
🛠️ Tools you might use:
Repo Reader / Parser · RAG Retriever · Image Generator · Web Search · Keyword Extractor · arXiv or Google Scholar Tool
You can organize agents however you like—by task, content type, or validation step—as long as they collaborate toward the shared goal.
Build a system that helps users explore patterns and trends across multiple AI/ML projects. The input is a list of GitHub repos or Ready Tensor publications, plus an optional user query (e.g., tool usage, evaluation methods, task types).
Support at least two of the following query patterns, each with multiple working examples:
🧠 Agents you might need:
Project Analyzer – inspects individual repos/publications for relevant attributes
Trend Aggregator – tallies, groups, or ranks signals across projects
Comparator – compares two project groups based on defined criteria
Summarizer – generates clear takeaways for the user
Fact-Checker – ensures outputs are grounded in project content🛠️ Tools you might use:
Repo Reader / Parser · Keyword/Tag Extractor · RAG Retriever · Web Search
Feel free to explore other collaborative use cases. Just make sure your system uses multiple agents to solve a shared problem that benefits from reasoning, tool use, or coordination beyond simple retrieval.
Your Mastering AI Agents project will be included in the review cycle for the month that matches your submission date.
To be assessed in that month’s cycle, submit your project by one of the following deadlines:
If you miss a deadline, submit before the next month’s due date to be reviewed in that cycle.
Each review cycle typically takes up to 2 weeks from the submission deadline, including feedback from the reviewer(s) and any updates or iterations you make in response.
To complete this project, you need to submit two deliverables:
Create a short publication on the Ready Tensor platform that:
The rubric is general-purpose. You still need to meet project-specific expectations (e.g. use of multiple agents, orchestration, etc).
Submit a repo that:
➡️ Next - Capstone Project Submission Guidelines
