ScholarMatch

🎓 ScholarMatch — Project Story

🌟 Inspiration

The spark behind ScholarMatch came from a simple, frustrating truth: students miss out on millions in available scholarships—not because they're unqualified, but because the search process is a nightmare.

I’ve watched brilliant students—especially from underserved backgrounds—believe college was out of reach, just because they didn’t know the right scholarships existed.

So I decided to build something for students, by a student. Something fast, intelligent, and actually helpful. Something that removes the guesswork from funding education. That’s how ScholarMatch was born.

---

🚀 What I Built

ScholarMatch is an AI-powered scholarship matching platform that takes a student’s profile and instantly ranks real opportunities based on a weighted scoring system.

The platform:

• Analyzes GPA, major, grade level, location, demographics, interests, and special circumstances
• Computes a personalized match score using a 7-factor weighted algorithm:

$$ \text{match_score} = \left( \frac{\sum_{i} w_i \cdot f_i}{\sum_{i} w_i} \right) \times 100 $$

• Provides analytics, deadline urgency, and a clean UI.
• Filters out scholarships the student is not eligible for.

All under 30 seconds, from profile → results → insights.

---

🛠️ How I Built It

Tech Stack

• Python for logic + data handling
• Streamlit for the UI
• Pandas for scoring + dataset operations
• JSON for persistent scholarship storage
• Plotly for interactive analytics

Core Matching Algorithm

Requirements are mapped to seven weighted factors:

Factor	Weight
GPA	20%
Major	25%
Grade Level	20%
Location	10%
Demographics	10%
Interests	10%
Special Circumstances	5%

Every factor contributes a weighted score, which is normalized to a final percentage.

---

🧠 What I Learned

1. Data design matters more than code

Creating a clean, consistent, real-world scholarship dataset was way harder than expected. I learned how crucial it is to structure data before coding anything.

2. Matching systems are more nuanced than I thought

I learned matching systems are complex, requiring careful handling of weights, normalization, fairness, and eligibility exclusions. It made me appreciate how recommendation engines work.

3. UX is everything

A technically good tool means nothing if users get lost. I learned the importance of color-coded deadlines and why analytics help users trust the results.

4. Shipping beats perfection

Deadlines force clarity. I learned to decide what was essential — and build that well.

---

⚔️ Challenges I Faced

1. Building a meaningful matching algorithm

Balancing weights for a “fair” score was tough. Too strict gives few matches, too lenient gives useless ones. Getting it right required extensive iteration.

2. Dataset integrity

Real-world scholarships have messy eligibility rules (fuzzy GPA cutoffs, vague demographics). Standardizing this data for automation took serious effort.

---

🌍 Impact & Future

ScholarMatch already works with 60+ U.S. scholarships, but my long-term goal is bigger:

Build the first open, global, crowdsourced scholarship database for Bangladesh + the Global South.

In future versions, I plan to add ML-powered match refinement, an AI essay assistant, and a full application management suite (user accounts, tracking, push reminders).

The dream is simple: make education accessible, one student at a time.

---

👥 Team Members

Solo Project Built by Tamzid Ahmed.

---

🤖 AI Tools Disclosure

I used GitHub Copilot during development to assist me with code generation for efficiency.

---

Thanks for reading — and happy shipping! 🚀

Built With

• json
• pandas
• python
• streamlit