Fourth-year Computer Science student at Dedan Kimathi University of Technology (DKUT), Nyeri, Kenya.
I build practical systems that solve real problems I've experienced firsthand — starting with the communication gaps in my own department.
class Ted Kimemia:
def __init__(self):
self.location = "Nyeri, Kenya 🇰🇪"
self.education = "BSc Computer Science @ DKUT (final year)"
self.languages = ["English", "Kiswahili", "Python", "SQL"]
self.mission = "Tech that works in low-resource environments"
def current_focus(self):
return "UniBot — shipping my final-year project 🚀"A WhatsApp-based academic communication and student engagement monitoring system for the DKUT Computer Science department — built to solve a communication gap I documented in my own department.
| Feature | |
|---|---|
| 📢 | Delivers lecturer & class-rep broadcasts to every student individually, with per-student delivery confirmation |
| 💬 | Answers student queries about timetables, deadlines & fees in English and Kiswahili/Sheng |
| Detects scheduling conflicts (venue, lecturer, year-group clashes) before they affect students | |
| 🧠 | Flags disengaged students for academic advisors using machine learning |
🔐 Security: SHA-256 hashing + Fernet encryption for student data protection at rest
✅ Shipped (click to expand)
- ✅ FastAPI backend — running, webhook live
- ✅ Full registration flow — JOIN → student ID → email → OTP → year
- ✅ Real DKUT timetable loaded (33 sessions)
- ✅ Student query engine — timetable / deadline / fee questions
- ✅ English + Kiswahili/Sheng language recognition
- ✅ Broadcast system with DELIVERED & READ receipts on real phones
- ✅ Per-student delivery confirmation + logging
- ✅ Conflict detection — validated against a real timetable clash
- ✅ ML Model 1 — intent classifier (F1 0.83, beats 55% baseline)
- ✅ ML Model 2 — disengagement detector (100% detection, 0 false positives)
- ✅ Phone number encryption at rest (Fernet + SHA-256)
- ✅ Live on real WhatsApp (Twilio)
- ✅ Real training data collection pipeline (Google Forms)
🚧 Roadmap
- ⬜ Cloud deployment on Railway (24/7 uptime)
- ⬜ React admin dashboard for lecturers
- ⬜ Production email OTP delivery
- ⬜ Dedicated Kenyan WhatsApp number
- ⬜ Retrain classifier on real student message data
- ⬜ Department pilot with a live class (4–8 weeks)
- ⬜ Final evaluation + load testing + report
Deploy → Kenyan number → Pilot with real class → Real ML results → Report
- 🤖 Machine learning applied to real-world problems
- ⚙️ Backend systems and API design
- 🌍 Building for the African context — multilingual, offline-tolerant, low-resource-friendly
- 🔐 Applied cryptography & data protection
"Technology should meet people where they already are — for my campus, that's WhatsApp."
— Building African solutions on African infrastructure
Solving real problems, one message at a time 💬🇰🇪