Machine Learning Engineer

Company Description

Chooch helps build and run the best Computer Vision AI applications that make cameras intelligent, enhancing productivity & revenues, while reducing operational costs. Chooch has been named a leading computer vision platform by IDC.

Role Description

We are looking for a Machine Learning Engineer (US-based) to join our team. This is a full-time, hybrid role with frequent hands-on experience with customer projects in the Bay Area. The Machine Learning Engineer will be responsible for the successful design, accuracy, and backbone behind our Computer Vision AI applications. This includes data preparation, training, configuring neural networks, deploying networks in real customer locations, and refining computer vision algorithms for healthcare, manufacturing, and retail customers. This role requires frequent interaction with technical and non-technical customers while working closely within our solutions engineering teams. The Machine Learning Engineering role has opportunity to grow, lead technical projects, and improve current processes while providing meaningful feedback to our product and solutions teams.

Product Description Chooch AI is a computer vision platform that helps hospitals automate supply room monitoring and inventory management. Using fixed cameras and AI, the platform continuously tracks supplies in storage areas and updates on-hand counts in real time. When inventory levels fall below defined thresholds, replenishment orders are automatically generated and sent to ERP systems, eliminating manual counts and barcode scanning. Hospitals using Chooch AI significantly reduce critical stockouts, optimize inventory levels, and improve supply chain staff productivity. By turning supply rooms into self-managing systems, Chooch AI enables clinical and supply chain teams to focus more time on patient care.

Responsibilities/What you'll do at Chooch:

• Design, train and fine-tune models for object detection, classification, segmentation, tracking and anomaly detection on real customer video and images.
• Adapt pretrained and foundation models (YOLO-family detectors, ViTs, CLIP-style and vision-language models) to new customer use cases with limited labeled data.
• Own data pipelines end to end: curation, labeling guidelines, augmentation, synthetic data, and dataset versioning.
• Optimize models for production and edge hardware using quantization, pruning, distillation, ONNX export and TensorRT, hitting agreed latency and accuracy targets.
• Build evaluation suites with clear metrics (mAP, precision/recall, false-alarm rates per camera-hour) and catch regressions before release.
• Deploy and monitor models in production, diagnose drift and failure cases, and close the loop with retraining.
• Work with product, solutions engineers and customers to turn vague requirements into measurable model specs.
• Write clean, tested, reviewed code and contribute to shared ML tooling.

Nice to have

• Experience with vision-language models (CLIP, LLaVA-style models) or open-vocabulary detection.
• Multi-object tracking and video analytics (re-identification, action recognition, temporal models).
• Few-shot learning, active learning, or synthetic data generation for low-data domains.
• MLOps tooling: MLflow or Weights & Biases, Kubernetes, Triton Inference Server, CI/CD for models.
• Domain experience in Healthcare, retail and supply chain, manufacturing inspection, workplace safety or video surveillance.

Key Components

• Model Development and Optimization: Design, develop, and optimize machine learning models for computer vision applications.
• Data Management: Collaborate with ML engineers to ensure the collection, processing, and maintenance of high-quality datasets necessary for model training and testing.
• Algorithm Implementation: Implement state-of-the-art computer vision algorithms to improve system accuracy and performance in real-world scenarios.
• Cross-Functional Collaboration: Work closely within the solutions engineering team to integrate machine learning models into the broader product and systems.
• On-Site Travel and Field Testing: Many of our projects are live deployments with major fortune 500 customers - testing new, innovative ML technologies. Traveling on site is sometimes needed to see the environment, test in the real-world, and ensure the technology is operating as expected.
• Performance Evaluation: Regularly evaluate the performance of computer vision models, using both qualitative and quantitative methods, and iterate to enhance accuracy and efficiency.
• Research and Innovation: Understanding of the latest developments in machine learning and computer vision, applying innovative approaches and technologies to solve complex challenges.
• Stakeholder Engagement: Collaborate with internal and external stakeholders to understand their needs and translate them into effective technical solutions.
• Problem-Solving: Address and troubleshoot complex problems that arise during the development and deployment of computer vision systems.

The ideal candidate should be deeply passionate about machine learning and its real-world applications, particularly in improving and leveraging state-of-the-art neural networks to make an impact.

Requirements

• 3-5 years of industry experience building ML systems, with at least 2 years focused on computer vision.
• BS or MS in Computer Science, Electrical Engineering, Math or a related field, or equivalent practical experience.
• Strong Python and hands-on experience with PyTorch (or TensorFlow) for training and debugging deep learning models.
• Shipped at least one computer vision model to production, and can explain the trade-offs made along the way.
• Solid grasp of CNN and transformer architectures for detection, segmentation and classification.
• Experience with OpenCV and image/video processing: decoding streams, preprocessing, augmentation.
• Working knowledge of model optimization for inference (ONNX, TensorRT, quantization or similar).
• Comfortable with Docker, Git, Linux and training on GPUs in a cloud environment (AWS/Azure).
• Sound ML fundamentals: loss functions, overfitting, class imbalance, evaluation metrics and experiment design.
• Clear written and verbal communication, including explaining model behavior to non-ML stakeholders.

Benefits

Chooch offers extensive benefits including health, dental, vision insurance, 401K with 3% match, extensive PTO, and life insurance policies included.