CampusHub

CampusHub is a digital platform designed to connect students, visitors, and universities. Its main goal is to make school and university guidance simpler and more accessible across the whole of DRC.

Comment 1

Inspiration

In the Democratic Republic of the Congo, students often struggle to find reliable information about universities, academic programs, tuition fees, campuses, and admission requirements. This information is usually scattered across social media, posters, outdated websites, and informal recommendations.

This problem inspired us to build CampusHub, a digital platform designed to connect students, visitors, and universities. Its main goal is to make school and university guidance simpler, more reliable, and more accessible across the DRC.

What it does

CampusHub allows users to:

  • search for universities and educational institutions;
  • explore campuses, faculties, programs, services, facilities, and admission requirements;
  • compare institutions according to location, tuition fees, and available programs;
  • create and manage student profiles and portfolios;
  • follow universities and academic publications;
  • like, comment on, save, and repost content;
  • discover opportunities published by institutions;
  • receive notifications and report inappropriate content.

The platform also includes CampusHub AI, an academic guidance assistant powered by GPT-5.6. Students can enter their preferred field, academic level, budget, and location. The assistant then searches the verified MySQL catalog and recommends suitable programs.

Students can also upload a report-card image for analysis. The extracted information is shown to the user for confirmation before it is used in the guidance process.

Universities have access to a separate institutional copilot that helps them prepare publication drafts, describe programs, explain admission requirements, and improve their institutional profile. The copilot never publishes automatically.

How we built it

CampusHub was built with three main components:

  • Frontend: React, Vite, React Router, and Lucide React;
  • Backend: Node.js, Express, JWT, Zod, and the OpenAI SDK;
  • Database: MySQL 8.

The backend follows this architecture:

Route → Middleware → Controller → Service → MySQL/OpenAI

Authentication and role-based authorization are handled with JWT and middleware. Zod validates user input, while Helmet, rate limiting, and centralized error handling improve API security.

GPT-5.6 is connected to strict internal tools that query only verified data stored in MySQL. This prevents the model from inventing universities, tuition fees, programs, or admission requirements.

We also created a demonstration mode that uses a deterministic MySQL recommendation engine when an OpenAI API key is unavailable.

Codex helped us review the architecture, design API routes, integrate OpenAI, create evaluation scenarios, fix technical issues, and verify the frontend build, backend syntax, and tests.

Challenges we ran into

One of the main challenges was preventing AI hallucinations. We wanted the assistant to reason about a student’s situation without inventing academic information. We solved this by allowing GPT-5.6 to use only data returned by controlled MySQL tools.

Another challenge was managing several user roles. Visitors, students, universities, and administrators require different permissions. We implemented JWT authentication, authorization middleware, and ownership checks to control access.

Maintaining data consistency was also difficult because CampusHub includes many connected resources, such as institutions, campuses, faculties, programs, publications, profiles, opportunities, notifications, and guidance files. We used MySQL constraints, transactions, triggers, stored procedures, and automatically generated codes.

We also needed to create a safe and reproducible demonstration without exposing real personal or institutional data. For this reason, we created clearly identified fictional demonstration data.

Accomplishments that we're proud of

We are proud of building an academic guidance assistant that does more than chat with students. It searches a verified catalog and recommends programs that match the student’s budget, field, level, and location.

We created five automated evaluation scenarios covering:

  • information technology in Goma with a limited budget;
  • health studies in South Kivu;
  • business and management studies;
  • mobility restricted to Bukavu;
  • a field or academic level that does not exist in the catalog.

All five scenarios were successfully validated. In the last scenario, the assistant correctly returns no result instead of inventing a program.

We are also proud of the clear separation between the student guidance assistant and the institutional copilot, as well as the ability to save recommendations in traceable guidance files.

What we learned

We learned that artificial intelligence becomes more useful when it is connected to reliable and controlled data.

We also learned how to:

  • integrate the OpenAI Responses API into an Express backend;
  • use tool calling to connect GPT-5.6 to MySQL;
  • process report-card images with a multimodal model;
  • design a platform with multiple user roles;
  • secure an API with JWT, Helmet, rate limiting, and Zod;
  • organize backend code using routes, middleware, controllers, and services;
  • create automated and reproducible evaluation scenarios;
  • provide a fallback mode when an external AI service is unavailable.

Most importantly, we learned that a trustworthy AI system should explain its recommendations, use traceable sources, recognize its limits, and never present uncertain information as fact.

What's next for CampusHub

Our next steps are to:

  • deploy a stable public demonstration;
  • integrate real and verified educational institutions;
  • expand the catalog of programs and admission requirements;
  • improve the academic guidance evaluation system;
  • optimize the platform for areas with limited internet connectivity;
  • allow institutions to regularly update their information;
  • measure the impact of recommendations on students’ academic decisions.

In the long term, CampusHub aims to become a trusted digital bridge between students and educational institutions in the Democratic Republic of the Congo.

Built With

Share this project:

Updates