AI & computer-vision strategy for imaging-heavy R&D
I help teams turn ambiguous scientific, technical, or product questions into tractable AI and computer-vision workflows: what to measure, what to automate, what to validate, and what not to build.
Scientist / R&D Leader / Company Builder
Applied Computer Vision & AI from research to deployment
Senior scientist, R&D leader, and company builder. Currently, I co-head Image & Data Analysis at ETH Zurich (ScopeM) and work with Harvard Medical School as an independent consultant.
Previously, I founded and led Tooploox Research as CEO, co-founded and served as CSO at the VR/game studio Cat-astrophe Games, and completed multiple exits, including Vratis, Tooploox Research, MobilWay, and a Bones Studio subsidiary. I help teams turn hard imaging and AI problems into workflows and products they can rely on.
My deepest domain is life sciences and microscopy imaging, with broader experience spanning robotics, AR/VR, interactive systems, and deep-tech R&D.
Most useful when a problem is too specialized for a generic ML team, too computational for a domain team, and too important to solve by blindly running the latest AI model.
Worked in labs at ETH Zurich, Harvard Medical School, Humboldt University Berlin, MIT, and INRIA.
Co-founded companies including Tooploox Research, Histopixel, and Cat-astrophe Games.
Advised Bones Studio, MobilWay, ZOWIE, Aeolus, and Clearlight Biotechnologies.
PI / grant execution EXAMODE H2020 (1M+ EUR), GameINN / City Stories (1M+ EUR), and Fortissimo 2 Microscopy in the Cloud (1M+ EUR).
Availability
Available for selected advisory and diligence engagements.
I take on a small number of selected fractional advisory and technical diligence engagements for imaging-heavy R&D teams, investors, and deep-tech companies.
Best fit: problems that are too domain-specific for a generic ML team, too computational for a domain team, and too important to solve by blindly running the latest AI model.
Leadership & companies
Executive and founder experience behind the technical work.
CEO & co-founder
MicroscopeIT
Computer-vision R&D company later merged into Tooploox Research; built and led a 30+ person R&D department.
Co-founder
Cat-astrophe Games
Interactive systems, game production, and AR/VR work through Cat-astrophe Games.
Angel investor
Deep-tech / life sciences AI
Early-stage investing and advisory around applied AI, robotics, life sciences technology, and product execution.
Founder & director
Histopixel
Advisory vehicle for scientific image analysis, applied AI, microscopy, and pharma-facing R&D.
Core capabilities
Applied Computer Vision & AI guidance from research to deployment.
Model validation and de-risking
I evaluate whether model outputs are scientifically and operationally trustworthy, including failure modes, quality controls, validation design, and decision criteria.
Scientific software and reproducible workflows
I prototype, review, and structure reproducible analysis workflows across microscopy, high-throughput imaging, and applied computer-vision systems.
Translation across science, ML, and business
I bridge domain experts, ML teams, product teams, investors, and grant consortia when no single group owns the whole problem.
Deepest domain: life sciences and microscopy imaging. Broader range: robotics, AR/VR, interactive systems, and deep-tech R&D.
Teaching
Teaching & community leadership
Teaching image analysis, AI-assisted scientific workflows, and scientific computing has been a recurring part of my work across ETH Zurich, ZIDAS, EMBL, and the bioimage-analysis community.
2022-now
Bioinformatics
ETH Zurich
2016-now
Zurich Image and Data Analysis School
SwissBIAS / ZIDAS
2020-2021
Deep Learning for Bioimage Analysis
EMBL Heidelberg
2016-2023
Microscopy Summer School
Center for Experimental and Clinical Imaging Technologies Zurich
2009-2012
Python, Matlab, modeling, and scientific programming
Humboldt University Berlin and earlier teaching
Ways to work together
Focused engagement formats for strategic direction, validation, and risk reduction.
Ongoing
Fractional R&D / AI advisor
Ongoing senior technical judgment for teams that need applied AI, computer-vision, or imaging R&D leadership, but do not need a full-time hire.
Investors
Technical due diligence
Independent assessment of AI, computer-vision, imaging, or scientific-software claims for investors, acquirers, and leadership teams: feasibility, data risk, model risk, team capability, and roadmap realism.
Scoped
Focused advisory review
A limited-scope senior review of an imaging-heavy AI or computer-vision project: what is technically credible, what is risky, what should be validated, and whether the next step should be build, adapt, buy, partner, or stop.
Background
Scientific, technical, and company-building history behind the current work.
Szymon Stoma works at the multidisciplinary interface of applied computer vision, artificial intelligence, and scientific R&D. His deepest specialization is life sciences and microscopy imaging, but the work spans advanced imaging, robotics, AR/VR, interactive systems, data interpretation, consulting, and startup R&D.
Brief Bio
A condensed path through science, software, and company building.
Academic / research
2024-now
Harvard Medical School
Independent Consultant, Boston, USA
2014-now
ETH Zurich
Co-head of Image and Data Analysis, Zurich, Switzerland
2011-2014
INRIA
Postdoc, Paris, France
2009-2011
Humboldt University Berlin
Postdoc, Berlin, Germany
2005-2008
University of Montpellier / INRIA
PhD, Montpellier, France
2000-2005
University of Wroclaw
MSc, Wroclaw, Poland
Companies / ventures
2016-now
Histopixel
CEO, Wroclaw, Poland
2017-2024
Cat-astrophe Games
Co-founder, Wroclaw, Poland
2013-2025
MicroscopeIT / Tooploox Research
Co-founder and CEO, Wroclaw, Poland
Current Focus
Turning scientific images into decisions, systems, and useful software.
Scientific imaging workflows
Co-head of Image and Data Analysis at the Scientific Center for Optical and Electron Microscopy at ETH Zurich, supporting microscopy-centered research with quantitative image analysis and data interpretation.
AI R&D and company building
Ex-CEO of Tooploox Research, previously MicroscopeIT, a software house specializing in R&D, AI, and Computer Vision. He led the merger with Tooploox and later headed a 30+ person R&D department.
Interdisciplinary R&D consulting
Independent consulting at Harvard Medical School on DARTS, an ARPA-H-funded antibiotic resistance program led by the Paulsson Lab, focused on image analysis and data workflows for high-throughput bacterial mother-machine experiments.
Institutions
Academic labs and companies where the work left a trace.
Academic
Founded Companies
Selected Activities
Projects, communities, investments, and shipped work.
2024
Don't Be Afraid 2
Co-producer of the game sequel, released on Steam by Eneida Games and Cat-astrophe Games.
2020-2026
BONES-SEED
Strategic R&D advisory work behind motion-data infrastructure later reflected in BONES-SEED, linked to the Bones Studio robotics data stack.
2020-2024
Essential Vision
Co-founder of Essential Vision, an open-source mixed-reality platform for teaching anatomy, simulation, and visual models.
2017-now
ZIDAS
Co-founder of an international school on image and data analysis for life scientists; see zidas.org.
2016-now
Bioimage analyst networks
Co-founding and community work across SwissBIAS, NEUBIAS, and GloBIAS, connecting analysts, training, and open resources.
2015-2019
CellStar
Co-founder of CellStar and the Evaluation Platform for long-term tracking of budding yeast cells.
2009-2020
STSE
Founder and main developer of STSE, the Spatio-Temporal Simulation Environment for biological modeling, image-derived geometry, and simulation.
2010-2018
Yeast Image Toolkit
Founder of the Yeast Image Toolkit, a software collection for yeast microscopy image analysis.
Selected Hands-On Examples
Things I modeled, simulated, or programmed directly.
INRIA / PhD
Growing meristem
Auxin transport, mechanics, and phyllotactic pattern formation in a simulated meristem.
HU Berlin / Postdoc
Aquaporin dynamics
Postdoc-era modeling work around membrane transport and plant-development mechanisms.
INRIA / Postdoc
Tissue reconstruction
Volumetric image analysis and reconstruction of complex biological tissue.
MIT / Visiting Scientist
GOL simulation
Multi-agent growth simulation connecting rule-based dynamics with image-analysis thinking.
MIT / Visiting Scientist
Yeast colony
Segmentation, tracking, and growth analysis for microscopy time-lapse data.
ETH / Staff Scientist
Essential Vision
Mixed-reality anatomy, simulation, and guided-learning work for HoloLens-style teaching environments.
Selected publications & active work
Recent and representative work across bioimage analysis, modeling, scientific software, and applied computer vision.
- BISCUIT: visual comparison of segmentation models in bioimage analysis Rantsiou E et al., Szymon Stoma last author, F1000Research, 2025
- 2024 OME-NGFF workflows hackathon Luthi J et al., BioHackrXiv, 2025
- Quantitative spatial analysis of hematopoiesis-regulating stromal cells Gomariz A et al., Nature Communications, 2018
- AutoTube: automated morphometric analysis of vascular networks Montoya J et al., Angiogenesis, 2018
- STL-based analysis of TRAIL-induced apoptosis Stoma S and Donze A et al., PLoS Computational Biology, 2013
- Flux-Based Transport Enhancement in Meristem Development Stoma S et al., PLoS Computational Biology, 2008