Table of Contents
Writings on software, startups, and building — each chapter a lesson learned
Designing for Delight: What Makes Developer UX Feel Magical
Breakdown of good developer UX: fast loads, subtle animation, feedback on interaction, component design clarity.
From UC Berkeley to Founder: What School Didn't Teach Me
Reflections on the gap between academic education and the real-world skills needed to build and grow a startup.
Engineering Popper's Infra: Firebase at Scale
A deep dive into the technical architecture behind Popper, including our approach to scalability and real-time synchronization.
Building in Public: Lessons from Launching Popper
Sharing the challenges and lessons learned while building Popper from the ground up and growing to over 1,000 users.
Building Love into Code: How I Made tomylovemiwa.com
The story of creating a Valentine's Day love game website with 6 interactive games and a surprise Japan trip planner.
Vibe-Coding a Website from Scratch, Less is More
What I learned building my personal website without a framework, a design system, or a plan — just a blank file, a feeling, and the discipline to stop before I went too far.
Curly-Hair-ai.com: Building a Domain-Specific AI for Haircare
The journey of creating an AI platform that helps users understand and care for their curly hair through personalized recommendations and routines.
Collaborating with BashNota: Bringing Revolutionary Computational Research Tools to the Masses
How I'm working with Taha Bouhsine, CEO of MLNomads and inventor of YAT, to bring BashNota - the world's greatest computational research tool - to researchers and developers worldwide.
The Gravity of Learning
Machine learning borrows its vocabulary from biology. The math underneath it is linear algebra. But the thing it might actually be describing is physics — specifically, gravity.
Small Models, Sharp Instincts
Before you can train big, you need to know how to read what your model is telling you. Optimizers, the batch size equation, and learning to diagnose training from the charts.
The Tokens You Don't See
Sequence packing, intra-document masking, and why the invisible data engineering of your training pipeline shapes what the model learns as surely as any architectural choice.
Infrastructure: The Unsung Hero
GPUs, memory hierarchies, NVLink, and PCIe — why the hardware layer underneath your training run shapes everything, and how to stop treating it as a black box.
Splitting the Work
From replicating models across GPUs to sharding every byte of memory — Data Parallelism, ZeRO, Tensor Parallelism, and Sequence Parallelism explained from first principles.
The Full Orchestra
Pipeline Parallelism, Context Parallelism, Expert Parallelism — the remaining three dimensions of distributed training, how all five compose, and the art of finding the right configuration.