8 years building and shipping backend systems and full-stack products, mostly solo or in small remote teams. I like owning things end to end, from infra to the API to the parts of the frontend that touch it.
Most recent project: <a href="https://myfinance.cc" rel="nofollow">https://myfinance.cc, a personal finance SaaS that parses M-Pesa and bank PDF statements. Transaction categorization runs on a Random Forest classifier (TF-IDF + temporal features, ~95% accuracy on live data); spend forecasting uses a per-user RF regressor with 31 engineered features; anomaly detection via Z-scores.
Also built <a href="https://watchdock.cc" rel="nofollow">https://watchdock.cc, an observability platform (error tracking with SHA-256 dedup, uptime/SSL monitoring, container health, an AI-generated daily ops summary posted to Slack). I also maintain a few async-first PyPI packages (kenyan-counties, daraja-sdk).
More projects and code on my site. Happy to talk about any of it.
NickLangat · · focus · HN ↗
Remote: Yes
Willing to relocate: No
Seeking: Full-time or long-term contract
Technologies: Python (Django, FastAPI), Go, PostgreSQL, MongoDB, Redis, Docker, Celery, Next.js, AWS
Résumé/CV: <a href="https://nicklangat.cc/" rel="nofollow">https://nicklangat.cc/
GitHub: <a href="https://www.github.com/nicksonlangat/" rel="nofollow">https://www.github.com/nicksonlangat/
LinkedIn: <a href="https://www.linkedin.com/in/nick-langat-26081119a/" rel="nofollow">https://www.linkedin.com/in/nick-langat-26081119a/
Email: nicksonlangat95@gmail.com
8 years building and shipping backend systems and full-stack products, mostly solo or in small remote teams. I like owning things end to end, from infra to the API to the parts of the frontend that touch it.
Most recent project: <a href="https://myfinance.cc" rel="nofollow">https://myfinance.cc, a personal finance SaaS that parses M-Pesa and bank PDF statements. Transaction categorization runs on a Random Forest classifier (TF-IDF + temporal features, ~95% accuracy on live data); spend forecasting uses a per-user RF regressor with 31 engineered features; anomaly detection via Z-scores.
Also built <a href="https://watchdock.cc" rel="nofollow">https://watchdock.cc, an observability platform (error tracking with SHA-256 dedup, uptime/SSL monitoring, container health, an AI-generated daily ops summary posted to Slack). I also maintain a few async-first PyPI packages (kenyan-counties, daraja-sdk).
More projects and code on my site. Happy to talk about any of it.