Hi, I am TJ
ML engineer, physics-trained. Two years shipping healthcare AI in production. MSc Computer Science at Lakehead, open to part-time and contract work.
Ad astra per aspera.
Selected work
Five entries: one production pipeline, three research studies, and the community work that keeps it honest. Open a log for the full record, or see all 28 on the work page.
About TJ
Until June 2026 I was a software developer at metricHEALTH Solutions in Barrie, Ontario, building FHIR-integrated data systems and production MLOps for clinical workflows. The team received the City of Barrie Mayor’s Award for Research and Innovation in 2024. I left to start an MSc in Computer Science at Lakehead University, where my research is on interventional auditing of ML models: pre-specified feature-removal tests that check whether a model is using the signal it claims to use.
A line I keep close is Ad astra per aspera, “a rough road leads to the stars.” The work I’m drawn to is the kind where the answer isn’t clean, where the system has to hold up under real conditions, and where you only know you got it right when the person using it tells you so. Physics and healthcare data both ask that of you.
My path started with a degree in physics and mathematics at the University of Mumbai, then two postgraduate certificates at Georgian College in AI Design and Implementation and Big Data Analytics. The clearest lesson came from three and a half years at the Barrie Public Library, where teaching emerging tools to the public shaped how I think about whether a system works for the person on the other side of the screen. I still sit on the Ontario Public Library Association’s Research and Evaluation Committee.
I’m looking for part-time or contract ML work alongside the MSc: roughly 20 to 24 hours a week during term, more between terms, remote or in the Barrie to Toronto corridor, starting now. If you’re building systems where reliability and human stakes both matter, email me or use the form below.
Where the spectra started: quantum mechanics, statistical mechanics, Michelson interferometry, and the habit of asking whether a result survives contact with real data.
AI Design and Implementation, then Big Data Analytics. Georgian Scholar and Dean's List in both cohorts.
Designed and taught a 24-session youth STEM program (Tech Titans, September 2025 to April 2026) and technology programs for the public: the clearest lesson in whether a system works for the person on the other side of the screen. Board member at Firebird Community Cycle, 2022 to March 2026.
Built FHIR-integrated healthcare data systems and production MLOps for clinical workflows. Team recipient of the 2024 City of Barrie Mayor's Award for Research and Innovation.
An interpretability audit of a LightGBM stellar spectral classifier on Gaia-ESO UVES data, and a paired-data study of structured FHIR versus LLM-narrative retrieval for clinical question answering.
Research focus: interventional auditing of ML models, the method behind the two manuscripts, now pointed at new classifier families. Coursework in research methodology and AI. Open to part-time and contract ML work alongside the program.
A system counts when it holds up in production under real conditions, not when the demo dataset behaves.
If a claim can carry a number, it does. 1,236 responses across Ontario public library systems: specifics are how trust is built.
You only know you got it right when the person using it tells you so.
"Tirthesh applies strong analytical thinking to a range of data analytics challenges and communicates findings clearly across teams."
Ruwindhu Hettige Don
Data Research Analyst, Georgian College
Send a transmission
Past the visible range now. Signal travel time from Barrie, Ontario: negligible. I read everything that arrives.