Physics-aware AI for thermal imaging.
ThermoLabs is built by a scientist and an operator — the two sides of the table that most thermography companies never bring together.
Two co-founders. Science and business.
Physics-aware AI needs rigorous research and disciplined execution. Most thermography companies have one or the other. ThermoLabs was built with both.
Homayoun Rezaie
Co-founder · Technical
I hold a PhD in thermal remote sensing, with a focus on city-scale heat loss measurement for energy efficiency analysis. My work bridges controlled scientific methods with complex real-world urban environments — where scale, material diversity, and environmental conditions introduce significant challenges.
At ThermoLabs, I lead the technical side: physics-aware machine learning pipelines, radiometric orthomosaic generation, and data acquisition across drone and satellite platforms. My focus is making thermal data as reliable and decision-grade as any other structured dataset.
Amirhossein Zahedi
Co-founder · Business
I lead the business side at ThermoLabs — product, pricing, go-to-market, and operations. My background is strategic product management: leading customer-facing products and cross-functional teams from discovery through delivery, with a bias for shipping and a track record of moving business metrics.
My focus at ThermoLabs is making rigorous thermography commercially scalable — turning services into products, pricing against markets (not hourly rates), and building operations that compound as the company grows.
Two gaps we're closing.
Thermography has forty years of history, but still a training problem and a data problem. Both are solvable — with physics-aware AI and a better business model.
No Canadian college or university teaches thermography.
Professional thermography certification costs $5,000+ and is delivered almost entirely abroad. Canada's energy auditors, inspectors, utilities, and engineering teams have no accessible, locally-priced training path — so the pipeline of qualified thermographers stays smaller than the demand.
Our answer: ThermoLabs Academy →Thermal imagery is captured, rarely decided on.
Reports sit in PDFs. Year-over-year inspections don't compound because last year's findings weren't structured for comparison. The bottleneck has always been manual analysis — which is exactly what physics-aware AI was built to solve.
Our answer: ThermoIntel and Services →Book a call with the founders.
Evaluating our services, exploring the Academy, or thinking about a partnership? We're happy to jump on a 30-minute call — no sales script.