Highlights
Block or report coderbenny
Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.
You must be logged in to block users.
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abuseFull-Stack & AI Engineer · Nairobi, Kenya 🇰🇪
I'm a Full-Stack & AI Engineer with 4+ years shipping software that real people use. I currently build telecom Value-Added Services and production LLM/RAG workflows at Zuri Health, where the systems I maintain support 50,000+ monthly healthcare transactions at 99.9% availability.
Outside of that, I independently design, build and ship products end-to-end - 10+ live applications across fintech, marketplaces and AI. I care less about frameworks-for-their-own-sake and more about the boring things that make software trustworthy: correctness under load, clean data and systems that don't wake you up at 3am.
- 🔭 Currently building: Snapit - an open-source, local-first encrypted clipboard vault (Flutter + Flask)
- 🧠 Currently learning: deeper MLOps & model serving, vector search at scale and Rust for systems work
- 🩺 Proudest ship: Glycosight - a clinical AI that grades diabetic retinopathy from retinal scans (EfficientNet-B4, sub-second inference)
- ✍️ I write about the hard parts of what I build - race conditions, LLM guardrails, zero-downtime deploys - on my blog
- 💬 Ask me about: fintech ledgers & M-PESA, production LLMs, telecom (SMS/USSD/IVR), or shipping solo
- 🌍 Based in Nairobi (GMT+3) · open to remote roles & collaborations
Languages
Frontend
Backend & AI
Data & Infra
| Project | What it does | Stack | Impact |
|---|---|---|---|
| Glycosight | Clinical AI grading diabetic retinopathy from fundus images | FastAPI · PyTorch · OpenCV · Next.js | Sub-second inference; EfficientNet-B4 trained on ~6k merged clinical images |
| Snapit | Local-first encrypted clipboard vault (open source) | Flutter · Dart · Flask · Redis | AES-256-GCM cross-device sync |
| SmartMinds | Digitizing Kenyan savings circles (Chamas) | Next.js · Flask · MySQL · M-PESA | 30+ active Chamas · 99.9% ledger accuracy |
| TabMemory | Accountability-first tab manager | Next.js · Flask · Redis/Celery | 1,000+ memories · 99.9% fulfillment |
| Gemify Africa | Discovery & booking marketplace | Next.js · Flask · GCP | 200+ listings · real-time booking |
| Errands By Us | On-demand errand marketplace | Next.js · Flask · MySQL | 500+ errands · 98% satisfaction |
More live products on my portfolio →
- Ship, then refine. I'd rather have something live and learning from real users than perfect on my machine.
- Money and health code is different. For ledgers and clinical systems I default to strong consistency, explicit locking and guardrails - "move fast and break things" doesn't apply when it's someone's savings or diagnosis.
- I write about what I build. Documenting the hard decisions makes me a better engineer and helps whoever hits the same wall next.
- Always learning in public. Every project teaches me something I didn't know when I started it - and I'm genuinely happy to be the least experienced person in the room if it means I grow.
I'm open to remote engineering roles, freelance builds and collaborations - especially anything in fintech, AI, or products for emerging markets. I reply to most messages within a day.
🌐 Portfolio · 💼 LinkedIn · ✉️ bhinnexclusive@gmail.com · 📝 Blog
Thanks for stopping by - if any of my work is useful to you, a ⭐ makes my day.
Pinned Loading
If the problem persists, check the GitHub status page or contact support.