Project file No. 06 · open since 2016 · status: rebuild in progress
Mad about cars. But logically, AI takes the wheel.
Not a portfolio: a workshop log. Mechanical engineering in Bangalore, analytics at RMIT, AI product work in Melbourne. Every entry below actually happened.
started with race cars. still just removing drag, honestly.† Project Coordinator at Health Careers International, Melbourne · Master of Analytics, RMIT · architected POMS, a placement system live in production · heading toward product management.
A national placement operation run on spreadsheets across two vendor systems. Shipped solo in 11 weeks, launched to near-zero adoption, then pivoted into an integration-fed system of record that refreshes itself every night.
6,417 students. One system of record.† role: Product Owner & sole builder.
Three GenAI and RAG systems built at Health Careers International. Two in production answering staff and student questions against the real internal record, plus a realtime voice assistant still waiting on its production slot.
2 in prod. 1 voice prototype.The big one. Credit-transfer assessments that once took a case officer 24–48 hours now resolve in minutes. Mapped the workflow first, then automated it end to end for 1,000+ users.
750+ hrs/yr back. ~$60K AUD.† role: Business Analyst, with substantial product ownership.
Five Power BI dashboards wired into decisions, not decoration. Built so the team reaches for evidence before instinct.
~60% more data-informed calls.IPC Health. A client-matching pipeline built from scratch: deduplication landing at 90% accuracy, record joins running 60% faster.
90% dedup. joins 60% faster.Mechanical engineering in Bangalore. Built formula-student race cars as vice-captain, published materials research across four papers, and trained with Mercedes-Benz R&D (MBRDI) through a joint coursework programme. Learned the rule I still work by: every system is a trade-off, and the data sheet doesn't lie for long.
An interview years back was where I first noticed Mercedes-Benz hiring computer-science grads, and I got curious why those skills were suddenly worth so much. COVID hit around the same time; my dad reckoned the data side was heating up. So I tested it properly: a year at Jigsaw Academy, then a Master of Analytics at RMIT here in Melbourne.
First road tests, at IPC Health. A client-matching pipeline built from scratch: deduplication at 90% accuracy, record joins running 60% faster. Learned to ship for the user, not the slide deck.
The full send, at Health Careers International. Led the Credit Transfer Automation platform end to end: assessments that took 24–48 hours now resolve in minutes, roughly 750 hours a year back, about $60K AUD. Shipped two GenAI & RAG systems to production (a third, realtime voice, is built and benched), wired five Power BI dashboards into everyday decisions, and picked up the Emerging AI Talent Award on the way.
learning the FujiA Fujifilm X-T30 III I'm still learning properly. Right now, deep in its colour science. Which, honestly, is the fun part.status: still learning.