Built in Montréal. Founded on Rigour.
Agmanic Vision is a Quebec-based specialist in AI-driven computer vision for industrial manufacturers. Founded 2019 by aerospace and robotics engineers.
Who We Are
Agmanic Vision is a Quebec-based specialist in AI-driven computer vision solutions for industrial manufacturers. We help product manufacturers enhance quality, improve functionality, and automate processes through vision-enabled technology that is engineered — not assembled. Our founders bring aerospace, nuclear, and industrial automation backgrounds from organizations like Rolls-Royce, Siemens, Symbotic Canada, and AMD. That pedigree means we approach every project with the rigour of systems engineering: failure modes are anticipated, tolerances are specified, and every deployment is validated against measurable acceptance criteria. We work directly with engineering and operations teams — no account managers in between. If we take on your project, Peter and Tristan are the engineers solving your problem.
- Founders hold P.Eng. and B.Eng. credentials — aerospace and robotics backgrounds
- Systems validated against client-defined acceptance thresholds before deployment
- Experience across aerospace, automotive, pharma, electronics, and warehouse automation
- Direct engineering engagement — no account management layer
Meet the Engineers
You work directly with the people who built the system — not an account manager.
Peter McLaughlin
Co-founder & Principal Engineer
P.Eng. with a background in aerospace systems engineering at Rolls-Royce and industrial automation at Symbotic Canada. Peter leads technical scoping, hardware design, and system integration on every project.
Tristan Pashley
Co-founder & AI/Vision Engineer
B.Eng. with experience in AI and computer vision systems at Siemens and AMD. Tristan leads model architecture, training pipelines, and embedded inference optimization.
Engineering First. Always.
We do not build vision systems from pre-configured platforms or shrink-wrapped software. Every system is scoped, designed, and validated from first principles. This means longer discovery phases, more rigorous testing, and explicit acceptance criteria agreed before a single line of model training code is written. It also means systems that work reliably in production — not just in the lab. We deliberately limit the number of projects we take on. We would rather do three things well than ten things adequately.
- Failure modes identified and mitigated during scoping — not discovered in production
- Acceptance criteria agreed in writing before development begins
- Classical and deep learning methods selected by what the problem requires, not by trend
- All source code, models, and training data are owned by the client upon project completion
Start a Conversation With Our Team
Tell us about your production challenge. We will tell you honestly whether vision technology is the right solution.