My Journey Into Neural Networks: From Confusion to Building
The honest, unfiltered story of how I went from staring at equations to building neural networks from scratch. Including tiny_cnn, The Residual Stream, and what's coming next.
Focused on geometric deep learning, ML systems, and the mathematical foundations of intelligent infrastructure. Building at the intersection of software, data, and community.
The honest, unfiltered story of how I went from staring at equations to building neural networks from scratch. Including tiny_cnn, The Residual Stream, and what's coming next.
Co-organising Agent Camp, winning 2nd place at Build With AI Makerere, attending Notion Community Day and Logos Circles Kampala. Real stories from Uganda's tech events.
Applying computational efficiency to reduce training cost and carbon footprint — low-resource ML, model compression, and energy-aware techniques.
From-scratch implementations of transformers, optimizers, and training loops using JAX and NumPy — focused on numerical stability and reproducibility.
Designing clear service logic with an emphasis on scalability, maintainability, and predictable system behaviour in real-world applications.
Turning complex outputs into understandable visuals and summaries that support both technical and non-technical decision-making.
Fast, gamified live quiz battles with solo, team, and league play. Esports-broadcast styled real-time arena with mobile-first tap-to-answer. Built with React, TypeScript, and Supabase Realtime.