Position Summary:
We are hiring a highly technical individual contributor to push the limits of our computer vision and machine learning capabilities. This is a high-impact, hands-on role for a research-minded engineer who wants to build and ship models. Much of the work involves large-scale visual understanding, extracting structured signals from imagery and reasoning about the real-world context behind a photograph, but we care more about deep ML/CV ability than any one problem area and welcome strong generalists.
Responsibilities:- Build, train, evaluate, and deploy computer vision and multimodal models, taking them from early prototype through to production
- Design systems that infer structured attributes and spatial context from imagery, combining learned models with geometric and heuristic reasoning
- Train and fine-tune models on large, diverse real-world image datasets, and build the pipelines to curate and label that data at scale
- Work with vision-language models (VLMs) and build rigorous evaluation frameworks to measure their accuracy on our tasks
- Develop and benchmark high-performance image retrieval capabilities with embedding models and vector indexing strategies
- Optimize models for inference latency and throughput using techniques like distillation, quantization, and GPU acceleration
- Read current research, prototype novel algorithms from academic literature, and turn promising ideas into reliable production code
- Implement efficient, scalable data pipelines and inference infrastructure
- Develop high-performance tooling in ML and data engineering
Requirements:- Experience building, training, evaluating, and deploying ML models in production
- Strong experience using PyTorch, JAX, or other deep learning frameworks to develop and optimize models
- Strong software engineering ability to build and maintain complex systems and work with large-scale datasets
- Ability to solve open-ended problems and quickly learn new domains
- Comfort operating with significant ownership and autonomy, making pragmatic trade-offs between model sophistication, velocity, inference and business constraintsBS, MS, or PhD in Computer Science or a related technical field, or equivalent practical experience
Nice to have:- Experience inferring structured, real-world attributes from images
- Experience training models on large-scale, real-world image datasets
- Familiarity with vision-language models (VLMs)
- Ability to digest academic literature, prototype novel algorithms, and bridge the gap between research and production code
- Experience building LLM or VLM pipelines and the evaluation frameworks to measure their performance
- Experience in an ML role at a growth-stage startup
- Publications in major ML or computer vision conferences (e.g., CVPR, ICML, ICCV, WACV)
Benefits:
Medical, Dental, Vision, STD and LTD PlansFSA - Medical and Dependent CareEAP and wellness programs13 Paid Holidays Unlimited PTOFlexible work environment - 100% remote 401(k) plan
