Beta
Vanderbilt students are looking for real problems to solve.
Students across AI 1010, the AI minor, the MSAI and the MSDS program are looking for research and applied projects to work on. Post a project here to connect with them.
Already posted a project? — who has applied, and where it stands.
Not on a class roster yet? Signing in will tell you who to ask.
Students from four programs are looking for projects
AI 1010, the AI minor, the MSAI and the MSDS program. Students in each are looking for projects with different goals and scope, so here is the kind of work that suits each of them.
Scoped, self-contained work with a clear finish line. These students arrive with a semester of AI fluency rather than years of programming, so what the problem is matters more than what it is built with. The largest group by far.
Projects with a real codebase and messier data behind them, where the shape of the problem isn’t obvious at the start and part of the work is figuring out what to build.
Hard research and engineering problems. Master’s students in AI, many with prior industry experience, looking for something substantial enough to defend as capstone work.
Data-heavy questions: modelling, forecasting, evaluation, and anything that needs a defensible answer and an honest write-up rather than a demo.
Ideas that make good projects
The best ones tend to be a task you already do by hand, or a question you have never had the time to chase down.
Your syllabus, readings and past exam questions become a private tutor your students can actually ask, instead of a PDF they skim the night before.
Open-ended survey responses, image sets, application piles — anything where the first pass is tedious and the judgement calls are the part that actually needs you.
You have the data and you have the questions. What’s missing is the thing in between that lets a colleague ask one without learning to code first.
A scoped, honest test of an idea you have wondered about, with a real write-up of where it falls down. A negative result a student can defend is a good project.
How it works
-
1
Describe your project
A short form: what the problem is, how many students it needs, and whether anyone may claim it or you would rather pick from applicants. You set a password so you can come back and see who is interested.
-
2
We read it
The Data Science Team at the Data Science Institute reviews every submission before students see it. If it needs shaping to work as a student project, we send it back with a note rather than quietly dropping it.
-
3
Students come to you
It goes on the board and students claim or apply. Sign back in whenever you like to see their year, major, experience and interests, and to accept the ones you want. Splitting a large project into sub-projects, each claimed separately, works well.
A reply within about a day when a student applies, and a willingness to meet them a few times over the semester. Students are graded on this work; a sponsor who goes quiet is the one thing that reliably sinks a project.
Scope it out with our project GPT, then post it
We built a custom GPT for exactly this. Describe the rough idea in plain words and it asks the questions we would ask — what the data is, what finished looks like, how much of a semester it is — then hands you back a file. Drop that into the submit form and every box fills itself in; you add your name and a password, read it through, and post it.
Find a project
Projects from faculty, labs, industry partners, and fellow students. Some you can claim outright — first come, first served. Others you apply for, and whoever posted the project picks.
The project board
Everything students can currently see, including your own. This is read-only — students are the ones who claim and apply — and it is here so you can see what your project sits alongside.
How to claim and work a project
Pick something that sounds interesting — not something you already know how to do. You aren't expected to arrive an expert. Learning it is the project, and everyone posting here knows they're working with students.
- Claim it. Or apply, if the person who posted it is choosing between students.
- Book your check-in. Find a time — do this the same day you claim.
- Agree when you'll next meet and put it in your calendar. Nobody is going to chase you.
- Keep notes as you go. What you tried, what worked, what you're stuck on.
- At every meeting, report out — where you are, and what you're doing next.
Your profile
Written once and sent with every application, so you're not retyping it each time. Whoever posted the project reads this — nobody else, and never the board.
Submit a project
Faculty, labs, industry partners, and students can post project ideas. Every submission is reviewed by the Data Science Team at the Data Science Institute before students see it.
Students in AI 1010 are graded individually, so a single project can't simply be handed to four people. If your project is bigger than one student, split it into sub-projects — distinct pieces that can each be taken and assessed on their own, while still adding up to the whole thing. Each sub-project appears as its own card with its own description, and can take one student or a small team.
What happens next
Your projects
Everything you've posted, how students are engaging with it, and — for apply-mode projects — who has applied.
Projects you've proposed
Projects you put forward for yourself. Once the Data Science Team at the Data Science Institute approves one it becomes yours automatically — there is nothing to claim, and nobody else can take it.
Review queue
Nothing reaches students until it's approved here. Approving acts on the whole submission, so a multi-part project never goes live half-approved.
Admin overview
Every project and claim across AI 1010, MSAI, and the minor — without reading through submissions one by one.
| Project | Source | Programs | Review | Status | Students | Runs | Off board | Posted |
|---|
Staff accounts
Everyone who can review submissions, manage the roster and export data. There is no self-signup for this — accounts are added here, by someone who already has one.
| Name | Last signed in |
|---|
What staff can do
Approve, reject and re-review every submission before students see it. Read the full roster and every claim. Export project and roster data. Reset anyone's password.
TAs get the same access — there is one staff role, not two.
Bug reports
Anything anyone told us was wrong, signed in or not. Open ones come first; resolving one keeps it, so the list stays a record of what was found rather than what is left.
Submitter accounts
Everyone who has posted a project: faculty, labs and industry partners. Accounts are created by the first submission, so there is nothing to add here — this is where you answer “I can't sign in” and keep names and organizations right.
| Name | Organization | Submissions | Last signed in |
|---|
Password resets
Submitters can reset their own password with the recovery code they were shown on first sign-in. If they've lost it, issue a temporary password here; the old code stops working and they get a new one when they next sign in.
Student roster
The list students pick their name from. Email is the only identifier stored — no VUnet ID, no grades.
| Name | Course | Section | Status |
|---|
