
Software Engineer at Microsoft · Full-Stack & AI Systems
Software Engineer at Microsoft, building production frontend systems at scale. I also built Zanira Clothing's full-stack e-commerce platform, with a custom AI conversational agent wired directly into a live product database. I work across TypeScript, Python, Next.js, and Django REST Framework, with a growing focus on AI integration and data science. Day to day I build agent-first: I delegate implementation to AI coding agents (Claude Code daily, Cursor occasionally, GitHub Copilot at Microsoft), then review, test, and accept or rework their output before it ships.
A dual-provider LangGraph ReAct agent (GPT-4o-mini by default, switchable to Claude via config) with 17 tools for inventory, orders, and payments, sharing a Postgres database directly with the Django backend, and serving customers over WhatsApp, web, and Instagram from one inbox.
A small-batch, Nairobi-made tailoring label. I built its full-stack e-commerce platform, handling the product catalog, capsule-collection drops, and order management, plus the AI customer assistant above.
Currently studying for an MSc in Data Science at the University of East London (via UNICAF), expected 2027. Coursework covers Geographic Information Systems and statistical analysis, using QGIS, R and Python (pandas) for spatial data visualization, statistical analysis and dataset cleaning.
Smaller builds and experiments alongside my main work.

A full-stack project I built before Zanira and continue to maintain. A luxury jewelry storefront covering the product catalog and customer-facing site, showcasing handcrafted, hypoallergenic, non-tarnish pieces with 18K gold coating.

A book-club platform built with React, TypeScript, and Django, with group creation, event scheduling, and member management for readers organizing recurring meetups.

A location-matching app for teachers built with React, TypeScript, and MUI. Matches educators by subject and grade level for reciprocal relocation swaps.

Kazi Outsourcing connects UK and US teams with Kenyan talent for virtual assistance, data and back-office work, and e-commerce operations. Landing page focused on clear pricing and how the managed pods work end-to-end.