§ Skills

A working list.

The tools I reach for first. Honest, not exhaustive, listed roughly in the order I use them.

Languages

  • Python primary, daily
  • JavaScript incl. TypeScript on RN
  • HTML & CSS
  • SQL

Web, backend

  • Django most production work
  • Django REST Framework
  • PostgreSQL · SQL
  • Auth, sessions, permissions
  • M-Pesa integration Daraja API
  • OOP, functions, data structures

Ops & observability

  • AWS deploy & host
  • Sentry error monitoring
  • Grafana dashboards & alerts
  • Git · GitHub

Web, frontend

  • Semantic HTML, modern CSS
  • Vanilla JS
  • Templated server-rendered UIs
  • Responsive, mobile-first layout

Mobile

  • React Native
  • Expo (Expo Go, EAS)
  • OpenGraphs
  • Android first, iOS to follow

Machine Learning

  • PyTorch
  • TensorFlow
  • Scikit-learn
  • Pandas · NumPy
  • Jupyter · PyCharm

ML, areas

  • Supervised & unsupervised learning
  • Convolutional neural networks
  • Recurrent / LSTM, attention
  • TCNs, dilated causal convs
  • Transformers (encoder)
  • Temporal / waveform prediction
  • Clinical signals & imaging

Data & research

  • PhysioNet credentialed access
  • Pediatric neurocritical-care waveforms
  • CHARIS (adult ICP)
  • CheXchoNet (chest X-ray + temporal)
  • Cleveland Clinic CV dataset
  • Leave-one-patient-out CV
  • CITI: Data/Specimens Only Research, COI

Tooling

  • VS Code · PyCharm
  • Jupyter Notebooks
  • Linux (Ubuntu) · macOS

Web3 (learning)

  • Solidity · smart contracts
  • Cyfrin Updraft course, 2024–25
  • Foundry basics

Soft

  • Technical writing & reports
  • Presenting research clearly
  • Working across time zones
  • Listening before designing

Languages (spoken)

  • English fluent
  • Swahili native
  • Korean beginner
§ A note on proficiency

I've left percentage bars off this page on purpose. They've never told anyone anything useful. A 90% next to a logo is a marketing claim, not an answer to "can you ship this thing".

A more honest version: I work in Python and Django every day. I'm comfortable shipping a React Native app to the Play Store. I'm in the middle phase with PyTorch, fluent enough to read papers and reproduce results, still learning the deeper craft of training and evaluating well. Sentry, Grafana, and AWS are part of the daily loop too. Solidity is newer; I worked through Cyfrin Updraft in the 2024–25 winter and am still in the "writing more than reading" phase.

If you want a project-by-project view of where each of these has actually been used, the projects page, research page, and résumé are the better read.