Sisonke

Inspiration

Our inspiration came directly from our shared experience as new Master's students at CMU-Africa. We quickly saw a major disconnect: while the academic rigor is world-class, it's so demanding that it’s easy to miss vital career-building activities.

As a team, we identified several core problems:

Fragmented Community Knowledge and 'Silos': It is difficult to know what our peers or even professors are working on. This means we are missing out on valuable collaborations and mentorship right next to us.
A Disconnect from Alumni: We know there is a huge network of successful alumni, but there is no clear or easy way to connect with them for mentorship or advice.
Unclear Career Paths: With so much coursework, many of us feel we don't have the time or guidance to even explore what we want to do post-graduation.
Inefficient Information Flow: WThe current communication method is impersonal and inefficient. This creates a high-noise, low-signal environment. For example, a student whose profile and goals are focused on Project Management, while already juggling a demanding course schedule, is constantly flooded with irrelevant notifications about IoT or Robotics events. This forces them to do the manual "sifting" work themselves, leading to notification fatigue where they eventually ignore all announcements and miss the critical opportunities that are relevant to them.
All this leads to Missed Opportunities and a feeling of disconnect.

Introducing Sisonke

What it does

Sisonke is an AI-powered, social-first career co-pilot designed specifically for the CMU-Africa community.

For Students: It uses AI to cut through the noise, delivering a personalized "Discover" feed of only the jobs, internships, and events relevant to their specific goals. It also helps them find and connect with peers, faculty, and alumni mentors using an AI-powered matching system.
For Alumni & Faculty: It provides a simple way to give back, offer mentorship, share opportunities, and find collaborators for their research.
For the Community: It breaks down silos by making projects, research, and interests visible. It also includes an AI Resume Coach and a Peer Review system to help students polish their applications and get immediate, actionable feedback.

How we built it

Given the 48-hour time constraint, we focused on building a high-fidelity, click-through prototype that would tell the complete story of Sisonke. We chose a modern, responsive, and fast stack that we could develop quickly in a collaborative environment.

Framework: Next.js 14 (using the App Router)
Language: TypeScript
Styling: Tailwind CSS
Host: Replit (for real-time team collaboration and development)

Our process was user-centric. We first mapped out a 10-screen user flow, starting with an onboarding process to capture a user's goals. We then hard-coded the prototype's data around a single "persona" (a student interested in "Ethical AI") to simulate a perfect, personalized AI-driven experience from end to end.

Challenges we ran into

Our biggest challenge as a team was scope. Our vision for Sisonke is huge, and it was difficult to decide which features to include in the prototype. We had to prioritize the core user story: Onboard -> Discover (Feed) -> Connect (Match) -> Improve (Resume).

The second challenge was simulating the AI. We couldn't build a real recommendation engine in 48 hours. We solved this by hard-coding the data, which let us demonstrate the value of the AI without having to build it. This was an effective storytelling technique that allowed us to present our full vision.

Accomplishments that we're proud of

We're also proud of our market research—we interviewed over 10 peers, and their enthusiastic feedback ("I need this now!") validated that we are solving a real, painful problem for our community. Finally, we're proud of our teamwork and how quickly we were able to learn, adapt, and build a complex product in Next.js under a tight deadline.

What we learned

We learned that the problem isn't a lack of opportunity at CMU-Africa, but a failure of personalization and connection.

We also learned that a good prototype is all about storytelling. By simulating some functionalities, we are able to communicate our full, ambitious vision far more effectively than if we had just built one small, working feature.

What's next for Sisonke

This prototype is our blueprint. Our team's next steps are clear:

Build the Backend: Implement a real database (like Firebase or Supabase) and user authentication.
Integrate a real LLM: Use an API (like Google's Gemini) to power the AI Resume Coach and to perform semantic analysis on user goals for real-time matching.
Launch a Pilot: Partner with the CMU-Africa administration to onboard a test group of students and alumni and start building the real Sisonke community. We believe this can fundamentally improve the student experience here.