Resources for Pre-Health & Pre-Med Advisors
Helping Pre-Med Students Build a Real AI Project — Not Just Another Extracurricular
STEAM in AI partners with pre-health advisors (WAAHP, NAAHP member institutions) to give premed and med-school-bound students a portfolio-ready AI project — built or researched 1:1 with an industry mentor, grounded in ethics-first design, and mapped to what admissions committees actually look for.
Two Tracks, Two Kinds of Value
Matching the Track to the Student's Story
AI Build — The Differentiator
Most premed applicants already have clinical volunteering and a standard research stint. Very few have built a working AI health tool. This track signals initiative and innovation precisely because it's rare in the applicant pool — a strong fit for students telling a "future physician-innovator" or digital-health story.
AI Research — Counts as Research Experience
Computational/AI research produces a poster, abstract, or paper — something advisors and admissions committees can place directly into the "research experience" bucket. It also doesn't require lab bench access, so it's a path to a real research credential for students without a nearby lab or PI connection.
We help you match the track to the student's story — not every premed student needs the same project.
AI Build
Student Project Examples
Examples shown, outcomes vary.
Final projects depend on readiness, interests, and mentor guidance.
MRI Knee Image Classification
JuliaBuilt an AI classification model for MRI knee images, evaluating accuracy and clinical limitations.
Accepted to Duke — Statistical Science / Data Science + CS
SkinSense — AI Skincare Analysis Tool
RaianBuilt an AI tool that analyzes skin health and conditions from images, with careful attention to accuracy limits and responsible use of visual health data.
Portfolio-ready AI health tool
AI Research — Forward-Looking Directions
Where Students Can Take AI Research in Health
These are proposed research directions — scoped and ready for a mentored student — not past student outcomes.
Health Equity Audit of a Diagnostic AI Model
Evaluate a public dermatology or radiology model's accuracy across different skin tones or demographic groups; document bias findings and clinical implications.
Disease Trend Forecasting Model
Build a predictive model using public health datasets (CDC, WHO, or similar open data) to forecast spread or risk of a chronic or infectious condition.
NLP Analysis of Patient Experience
Text-mine patient forums, clinical notes, or medical literature to surface unmet needs, treatment barriers, or communication gaps in a specific condition community.
Each direction pairs a real public dataset with mentor guidance, and is scoped to produce a submittable poster, abstract, or paper.
Ethics-First by Default
Why This Fits an Ethics-First Framework
Every project — Build or Research — includes bias, safety, and human-oversight evaluation as part of the process, not as an add-on.
For health specifically, this maps directly to how medical AI is actually scrutinized in practice — clinical validation, equity of diagnostic accuracy across populations, and appropriate human oversight in decision-making.
For Advisors
How This Works
STEAM in AI works alongside pre-health advisors the same way we work alongside independent educational consultants — we don't replace your advising relationship with the student.
We provide the specialized AI mentorship layer: project design, 1:1 industry mentorship, and a portfolio-ready outcome your student can speak to in interviews and applications.
Questions? Reach out to Shilpi at shilpi@steaminai.org
Schedule a Conversation