Leszek W.-AI/ML Evaluation & Infrastructure Expert

If you are a CTO or an engineering leader whose AI products are held back by:

• late regressions
• unclear quality metrics
• costly validation workflows

I help establish repeatable evaluations and infrastructure that support engineering and release decisions, minimizing the risk of late-stage defects and avoidable delays. My work helps teams evaluate algorithm quality, release products, and reduce infrastructure costs. I have a 15-year track record of excellence at various tech companies:

• Providing CV and ML teams at Magic Leap with a highly scalable and robust verification solution, enabling the release of ML1 and ML2 products.
• Delivering an agentic AI system as a contractor for Google XR, allowing Android XR teams to leverage algorithm quality and performance insights bolstering data-driven product decision making.
• Restructuring cloud verification platform containing 1PB of data for cost-efficiency, saving annually USD 1 million.
• Founding member at Archilyse AG, a PropTech startup that raised USD 5.3 million funding.

My leadership capabilities are confirmed by:

• Growing a multidisciplinary engineering team at Magic Leap up to 20 members consisting of computer vision researchers, machine learning and cloud engineers.
• Driving Magic Leap and Google synthetic data team initiatives and creating joint strategies as part of the companies' strategic partnership.