Build Research Teach

Hi, I'mJC Diamante

I build systems, research practical AI, and teach what I learn by turning difficult technology into something people can use.

View My Work

JC Diamante smiling in front of an AWS sign

Primary practiceSoftware Engineer
AI Strategist
OSS Contributor
ML Researcher
Tech Educator

Hackathons

I love collaborating, building, and pitching solutions to real-world problems.

View 8 selected builds →

Studying

Machine Learning

MSAI @ UT Austin 🤘

Checking your timezone…

Connecting to Manila…

Philippine time

Loading…

Certified ✓

8 professional · AWS · Azure · GitHub · NASA

Browse all credentials →

Research

Computer vision for Philippine agriculture and waterways.

02 peer-reviewed papers

My setup

Most of my setup is boring on purpose.

Java and Postgres do the work. Newer things earn their place slowly.

See what’s on my desk →

01What I write in

MergedRust → Zed #61698

02Where it runs

At work30-odd microservices

03AI I actually use

BoundarySensitive data stays local

04The desk

OwnedM5 Mac · Pi 4B

Zed · Neovim · tmux · Git

M5 Mac · Raspberry Pi 4B

Selected writing

Notes from the work.

Featured writing

Jul 26, 2026Open Source8 min read

Your First Open-Source Pull Request: What I Wish I Knew Before Mine

My first merged pull request to BetterGov PH changed where I put my energy. This beginner-friendly guide explains how to choose an issue, use AI responsibly, open a PR, and ask for help.

Read featured article

Jul 21, 2026Machine Learning10 min read

Five Builders, One AMD MI300X: How We Built Ani for the AMD Hackathon

Read article

Jul 7, 2026Local AI10 min read

Can't Use Cloud AI on Patient Data? Here's How I Built an ICU Mortality Predictor with Local LLMs

Read article

Keep reading beyond the selected three.

Software, applied AI, research, and the tools behind the work.

Browse every article

Build · Research · Teach

My Journey

One question at a time

It started with a broken website at eight years old and a stubborn need to build the next one anyway. Every chapter since has been another question nobody assigned.

Chapter 012011

Nobody asked the question and nobody was going to answer it. Sixteen years later this is still the whole method.

Chapter 022018 – 2020

Cloud load balancers and a Lego robot, both under a clock, both in high school. The first taste of trading cost against performance on a stage.

Chapter 032020 – 2023

A national data-science scholarship, satellite imagery, a self-driving car in simulation, two papers, and a thesis — all of it the same question at different resolutions.

Chapter 042022 – 2023

Art pipeline, minting product, community, and eight people to coordinate. The answer turned out to be yes, and the market still had the last word.

Chapter 052024 – 2025

Cadetship into production ownership. The answers were mostly unglamorous: consumer offsets, stale caches, container configs that drift between environments.

Chapter 062024 – 2026

Wildfires from aerial imagery, with Arizona State. A model that scores well on a clean dataset can still fail the people waiting on the warning.

Chapter 072025 – now

Agent skills, merged upstream contributions, and clinical models that never see the network. The current question, still unassigned.

Hi, I'mJC Diamante

I build reliable systems and better ways to work.

Hackathons

Machine Learning

Connecting to Manila…

Certified ✓

Research

Most of my setup is boring on purpose.

A few things I’m glad I helped build.

Notes from the work.

Your First Open-Source Pull Request: What I Wish I Knew Before Mine

Five Builders, One AMD MI300X: How We Built Ani for the AMD Hackathon

Can't Use Cloud AI on Patient Data? Here's How I Built an ICU Mortality Predictor with Local LLMs

My Journey

“Why is my website broken?”

“Can I make the machine do it?”

“What can I learn from the data?”

“Can I build the entire thing myself?”

“What breaks at 94 million users?”

“Can a model see it before we do?”

“How much of this can I safely automate?”