Tartan OppHub

AI - Powered solution that streamlines career opportunity discovery and engagement for students and career offices using automation and personalization

Comment

Inspiration

At CMU-Africa, we observed a critical problem affecting both students and career services staff.
Career and growth opportunities were scattered across countless platforms — email lists, Slack channels, LinkedIn posts, and various job boards.
Students spent hours searching through this digital chaos, often missing crucial deadlines.

Career service staff manually curated opportunities from multiple sources, creating duplicated effort and inefficiency.
There was no centralized system to track applications, measure engagement, or provide personalized recommendations.
We saw the stress this caused our community and knew there had to be a better way.

What It Does

Tartan OppHub is an AI-powered platform that serves as the central hub for opportunities aligned with CMU-Africa students.

For Students, it provides:

  1. AI-Powered Matching: Uses NLP and machine learning to match students with relevant internships, scholarships, and research opportunities based on their skills and interests.
  2. One-Click Applications: Automatically generates personalized application emails with attached CVs.
  3. Application Tracking: Manages the entire application journey from saved to applied, with reminders for applications marked as important by students.

For Career Services Staff, it offers:

  • Automated Opportunity Aggregation: Syncs opportunities from multiple trusted sources daily, weekly, bi-weekly, and monthly.
  • Analytics Dashboard: Tracks student engagement and success metrics.
  • Centralized Management: Streamlines opportunity curation and distribution.

How We Built It

We architected Tartan OppHub as a modern, scalable web application with a microservices approach.

Frontend Architecture

  • Responsive and modern UI (Tailwind CSS, React)
  • Context API for state management across student and admin dashboards
  • React Router for role-based navigation and protected routes

Backend Services

  • Python Flask microservice for AI matching using Sentence-BERT embeddings
  • Role-based access control for students and staff
  • RESTful API design with robust error handling and JWT authentication

AI & Matching Engine

  • Natural Language Processing using Sentence-BERT for semantic similarity matching
  • Resume parsing and skill extraction from uploaded CVs
  • Cosine similarity algorithms for opportunity-student matching:
    $$ \text{similarity}(A, B) = \frac{A \cdot B}{|A||B|} $$
  • Personalized email generation using template-based systems

Data Management

  • JSON-based data layer for rapid prototyping
  • Structured data models for users, opportunities, and applications
  • Analytics tracking for engagement metrics

Challenges We Faced

Technical Challenges

  • Implementing accurate AI matching that considers both explicit skills and implicit interests
  • Creating a seamless onboarding flow that captures comprehensive student profiles
  • Curating datasets for opportunity-student matching
  • Building an email generation system that creates personalized, professional applications

UX/Design Challenges

  • Creating intuitive interfaces for both technical students and non-technical staff
  • Designing a multi-step onboarding process that feels engaging rather than burdensome
  • Balancing feature richness with simplicity in both student and admin dashboards

Integration Challenges

  • Simulating external API integrations (LinkedIn, Indeed, etc.) within prototype constraints
  • Implementing real-time synchronization features for opportunity aggregation
  • Creating analytics that provide meaningful insights without overwhelming users

Accomplishments We're Proud Of

  • Built a dual-sided platform that serves both students and career services with tailored experiences
  • Implemented sophisticated NLP matching that provides genuine value to students
  • Delivered a complete user journey from profile creation to application tracking
  • Developed a production-ready prototype with comprehensive features and polished UX
  • Incorporated CMU-Africa identity and community values throughout the platform

What We Learned

Technical Insights

  • The importance of modular architecture when building multi-role applications
  • How to effectively implement AI/ML features in a web application context
  • Best practices for authentication and authorization in educational platforms
  • Techniques for handling complex user onboarding flows

Product Insights

  • The critical balance between automation and user control in AI-powered systems
  • How to design for two distinct user personas with different needs and technical abilities
  • The value of progressive disclosure in complex feature sets
  • Importance of building trust when handling sensitive career data

What's Next for Tartan OppHub

Short-Term Goals

  • Deploy pilot program with CMU-Africa Career Services
  • Integrate with real external APIs for opportunity aggregation
  • Implement real email-sending capabilities
  • Add mobile app version for on-the-go access

Medium-Term Vision

  • Expand to other universities in the CMU network
  • Implement advanced analytics and predictive placement features

Long-Term Aspirations

  • Become the standard career services platform for African universities
  • Build employer partnerships for direct opportunity posting
  • Develop AI-powered career path recommendations

Built With

+ 75 more
Share this project:

Updates