1 day ago
Base Salary
$200k - $260k/yr
Responsibilities
- Design computer vision architectures for fine-grained item identification using foundation vision models adapted to the marketplace catalog.
- Set category-specific accuracy standards and build calibrated confidence scoring to communicate prediction uncertainty.
- Develop model-error feedback loops to prioritize future labeling with the labeling team.
- Balance category coverage expansion against accuracy improvements in existing categories.
- Own the transition from model development to production serving while managing latency, cost, and reliability.
- Explain model capabilities, limitations, and technical tradeoffs to technical and non-technical stakeholders.
Requirements
- At least 5 years of applied computer vision experience, including shipping a production system at meaningful scale.
- Hands-on experience with fine-grained or instance-level classification involving visually similar items.
- Strong fluency in PyTorch or TensorFlow and practical experience fine-tuning and deploying vision transformers or CNNs.
- Experience designing and running vision model evaluations beyond loss metrics.
- Experience owning model serving decisions involving latency, cost, reliability, and calibrated confidence scoring.
- Ability to take technical ownership of open-ended problems and communicate tradeoffs clearly.
- Experience with active learning, human-in-the-loop labeling, low-latency model APIs, or an early-stage startup is a plus.
Benefits
- Annual base salary of $200,000 to $260,000.
- On-site role in San Francisco, California, United States.
Tech Stack
PyTorchTensorFlow