I'm Ramadan, a petroleum engineer at an oil and gas operator in Indonesia. I read wells, data and my own training the same way: collect what actually happened, find the pattern, then act on it. This page is the work that comes out of that habit.
Production engineering taught me to trust data over opinion. Building tools made that faster. Endurance training is where I test it on myself every morning.
Production engineering on mature onshore fields.
Tools that turn scattered records into decisions.
Marathon runner, training for triathlon in 2027.
Built in the evenings, used every day. None of them shows data from my employer.
The distances and personal bests on the front page are not typed in. After every Garmin sync, my own system adds up the week and the year, checks the records against Garmin and publishes them to this website.
A personal AI system. One orchestrator routes each message to a specialist agent; PostgreSQL with pgvector is its memory, Telegram is how I talk to it and Notion is its dashboard. It writes my morning brief, plans training from Garmin data, coaches my English practice and keeps my documents searchable by meaning.
A browser tool that masks spreadsheets before they are shared: names become codes that still join across tables, numbers are scaled so trends survive, dates move together and free text is cleaned. Nothing is uploaded; the page has no network access at all.
| well | date | oil, bbl/d | remarks |
|---|---|---|---|
| ALPHA-02 | 2026-09-01 | 159.0 | Waiting on rig for ALPHA-02 |
| WELL-KNRQK3VR | 2025-08-10 | 173.26 | Waiting on rig for WELL-KNRQK3VR |
Synthetic demo data: the first row before masking, the second after.
A team of subsurface AI agents that work one well together, the way a geoscience and engineering team would, with an engineer reviewing every result.
petronome.id