<aside> 🗣

Interested in this role? Email us at [email protected]

</aside>

About Vardera Labs

Vardera Labs is building the next generation of infrastructure for the art, auction, and collectibles industry—combining cutting-edge AI, computer vision, and automation to power catalog creation, data strategy, and operational excellence for auction houses worldwide. We’re a small, venture-backed team with a bold mission: to modernize a multi-billion-dollar market at a critical moment of generational ownership transfer and digital transformation.

What You’ll Do

• Design and deploy computer vision systems that identify, grade, and authenticate physical assets across categories like coins, comics, jewelry, and fine art.
• Build and fine-tune multimodal models that combine visual, textual, and structured data to deliver accurate, explainable outputs.
• Prototype rapidly: leverage foundation models, open-source frameworks, and custom training pipelines to reach production-ready performance quickly.
• Define and evolve accuracy and quality metrics beyond standard loss functions — optimizing for the real-world economics of error and trust.
• Create robust evaluation and monitoring pipelines that track performance by category, lighting condition, and visual defect type.
• Work directly with founders and customers to turn ambiguous problems into measurable, repeatable, high-accuracy solutions.
• Hustle on speed: ship usable models in days/weeks, not months, while stacking learnings into a long-term defensible data moat.

What We’re Looking For

• Strong foundation in computer vision, deep learning, and applied ML (e.g., segmentation, detection, embedding models).
• Experience with modern CV toolkits — PyTorch, YOLO, SAM, Detectron, CLIP, or similar.
• Comfort working across image data pipelines, from annotation strategy to model deployment and feedback loops.
• Pragmatic engineer who balances rigor with velocity: knows when to experiment, when to productionize, and when “good enough” drives impact.