Introduction

The capstone project is the final requirement for graduation from the AI/ML bootcamp. This is your opportunity to demonstrate the skills and knowledge you have developed throughout the program by building an end-to-end machine learning project.

Your capstone serves two purposes:

  1. Graduation Requirement: Successful completion and presentation of your project is required to complete the bootcamp.

  2. Portfolio Piece: This project will become a key component of your professional portfolio, showcasing your abilities to potential employers.

Think of this as your chance to prove you can take a problem from concept to solution using the tools and techniques you have learned.

What is an end-to-end machine learning project?

An end-to-end project demonstrates your ability to work through the complete machine learning lifecycle:

  • Problem Definition: Identify a meaningful question or problem to solve
  • Data Acquisition: Gather, access, or generate appropriate data
  • Data Exploration: Understand your data through analysis and visualization
  • Data Preparation: Clean, transform, and prepare data for modeling
  • Model Development: Select, train, and evaluate appropriate models
  • Iteration: Refine your approach based on results
  • Communication: Present findings and demonstrate your solution

You do not need to build a production-ready system. The goal is to show competency across these stages and document your process clearly.

Project options

You have flexibility in choosing your project. Options include:

  1. Original Project: Develop your own idea from scratch. This is encouraged for students who have a specific interest or problem they want to explore.

  2. Default Capstone: Choose from one of the provided capstone options. Even if you select a default project, you will still work through the full project lifecycle including planning, proposal, and roadmap phases.

Regardless of which option you choose, the planning and documentation process is the same.

Project phases

Your capstone will progress through the following phases:

Phase 1: Concept pitch (before official start)

You will present a brief elevator pitch of your project idea to a small group of peers. This is an informal opportunity to share your thinking, get early feedback, and refine your concept.

Phase 2: Proposal (before official start)

Submit a written proposal that clearly defines what you will build. This includes your problem statement, data sources, approach, and expected deliverables. For projects involving model training, data verification is required at this stage.

Phase 3: Roadmap (before official start)

Create a detailed timeline with weekly milestones, tasks, and deliverables. Define your Minimum Viable Product (MVP) and any stretch goals. For team projects, this is where you assign responsibilities.

Phase 4: Build (official project period)

Execute your plan. Work through your roadmap, track progress, and adapt as needed. Regular check-ins will help keep you on track.

Phase 5: Presentation and demo (end of project period)

Present your completed project to your peers and instructors. This includes demonstrating your solution and explaining your process, decisions, and findings.

Expectations

To successfully complete the capstone, you must:

  1. Complete all planning phases (pitch, proposal, roadmap)
  2. Deliver a working project that meets your defined MVP
  3. Maintain documentation of your process and decisions
  4. Present and demo your project to the cohort

Quality matters more than complexity. A well-executed simple project is better than an incomplete ambitious one.

Project management

You are expected to use professional project management practices throughout your capstone. This includes:

  • Version control using Git and GitHub
  • Tracking tasks and progress using GitHub Projects or a similar tool
  • Regular commits with meaningful messages
  • Clear documentation

Using GitHub Projects is strongly recommended. This integrates your proposal, roadmap, and code in one place and demonstrates professional workflow practices that employers value. Even if your final deliverable is a single notebook, managing your project professionally will set you apart.

If you have a specific reason you cannot use GitHub, discuss alternatives with your instructor. This should be the exception, not the rule.

Team projects

Team projects are permitted with instructor approval. If working in a team:

  • Maximum team size is typically 3-4 members
  • All team members must contribute meaningfully
  • Roles and responsibilities must be clearly defined in your roadmap
  • Each team member should be able to explain all aspects of the project
  • Collaboration and delegation should be documented

Deliverables

At minimum, your project should include:

  1. Completed proposal document
  2. Project roadmap with milestones
  3. Working code (notebook, scripts, or application)
  4. Documentation explaining your approach and findings
  5. Final presentation

Your specific deliverables will depend on your project and should be defined in your proposal.

Presentation requirements

At the end of the project period, you will present your work to the cohort. Your presentation should:

  • Be approximately 10-15 minutes (specific time will be confirmed)
  • Explain the problem you addressed and why it matters
  • Describe your data and approach
  • Demonstrate your solution
  • Discuss results, challenges, and what you learned
  • Be prepared for questions from peers and instructors

This is a professional presentation. Treat it as practice for presenting to stakeholders or in a job interview.

Timeline overview

Before April 1 (pre-work)

  • Complete concept pitch
  • Submit project proposal
  • Create project roadmap

April 1 - project end date

  • Execute your plan
  • Regular check-ins and progress updates
  • Adapt as needed

Final week

  • Complete project
  • Prepare presentation
  • Present and demo to cohort

Getting help

You are not alone in this process. Resources available to you:

  • Instructor and TA office hours
  • Peer feedback during pitch sessions
  • Regular course meetings
  • Cohort collaboration (while maintaining academic integrity)

Ask for help early if you are stuck. It is better to adjust your plan than to fall behind.

Academic integrity

Your capstone must represent your own work. You may:

  • Use libraries, frameworks, and tools as intended
  • Reference documentation and tutorials
  • Get feedback from peers and instructors
  • Use code snippets with proper attribution

You may not:

  • Submit work completed by others as your own
  • Copy code without attribution
  • Use AI tools to generate your entire project without understanding

When in doubt, ask your instructor.

Final notes

The capstone is designed to be challenging but achievable. The planning phases before the official start date may feel like extra work, but investing time upfront will make your build period much more productive.

This is your project. Choose something that interests you, plan carefully, and execute with intention. The skills you demonstrate here, both technical and professional, will serve you well as you move into your career.

Good luck, and have fun with it.