Nairobi, Kenya · Software Developer
Five years building enterprise integrations, academic platforms, and data pipelines across Kenya. Currently researching NLP-driven helpdesk automation for my MSc — and preserving endangered languages through technology.
about
I spend most of my time designing and maintaining the integrations that quietly hold enterprise systems together — the kind of work where you have to understand the business, the data, and the code all at once.
My background spans HR platforms, academic management systems, and government-facing regulatory portals — building from requirements gathering with non-technical stakeholders all the way through to architecture, deployment, and documentation.
Outside of systems work, I run a personal platform dedicated to language and cultural preservation — building digital resources, community engagement tools, and educational systems — because technology is one of the most powerful tools to keep languages alive.
Current focus — ML & NLP
I'm actively moving into machine learning — specifically natural language processing applied to information systems. My MSc research involves building a model to automatically classify and route helpdesk queries, using real production ticket data (~30,000 tickets).
I'm interested in where NLP meets the messy realities of enterprise data — not clean benchmark datasets, but the noisy, inconsistent stuff from real systems.
technical skills
work experience
Software Developer (Systems Developer)
Software Developer
notable projects
Integration
HR System Integration Bridge
Custom PHP/Laravel bridge connecting two enterprise HR platforms via REST APIs. Automates data synchronisation for 1,500+ staff records, replacing a labour-intensive manual reconciliation process.
Enterprise
Academic Management System
Java / Spring Boot / Oracle DB system managing student and academic records for ~6,000 users at a major Kenyan university. Developer and ongoing system administrator.
ML / NLP Research
NLP Query Routing System
MSc research project: building a model to automatically classify and route student helpdesk tickets using real production data (~30,000 tickets). Compares Logistic Regression, SVM, and BERT.
Personal
Language Preservation Platform
A cultural preservation platform for endangered languages — public learning system, community resources, and contributor portal. Dedicated to maintaining and documenting indigenous language collections.
ML Pipeline
Loan Default Prediction Pipeline
End-to-end ML pipeline covering EDA, hypothesis testing, feature engineering, and multi-model comparison. Achieved AUC 0.940 with Gradient Boosting on a structured finance dataset.
Government
Regulatory Data Portals
Production government-facing portals for regulatory data submission and licence management. Required strict auditability, error handling, and data integrity standards throughout.
education
MSc. Information Technology
Business Intelligence & Data Analytics
Personal Leadership Programme
Team leadership & stakeholder engagement
BSc. Computer Science