Senior ML Engineer

The company is a leading provider of vehicle lifecycle solutions headquartered in Chicago. It supports companies involved in building, insuring, and replacing vehicles by delivering advanced mobile, artificial intelligence, and car technologies. The platform connects a network of over 350 insurance companies, 24,000 repair facilities, hundreds of parts suppliers, and multiple third-party data and service providers to enhance productivity and improve end consumer experiences.

• Advanced knowledge of Python (native, Pandas, ScikitLearn, Tensorflow or Pytorch, PyStats, Pydantic)
• Experience with AWS tools for ML Engineering and ML deployment (Sagemaker, Lambda, Cloudformation/CDK, Step Functions)
• Advanced knowledge of SQL and Data Modeling
• Experience with GenAI for document intelligence, including prompt engineering, RAG (Retrieval Augmented Generation), multi-modal models (vision + text), and production deployment using AWS Bedrock or Azure OpenAI APIs
• Experience in experiment design (power analysis and hypothesis testing)
• Proficiency in written and verbal communication for remote and asynchronous work
• Ability to clearly communicate complex technical problems and propose iterative solutions
• Experience owning a feature from concept to production, including proposal, discussion, and execution

Nice to have:

• Experience with document processing tools (AWS Textract, Azure, Document Intelligence, or similar OCR/layout detection systems)
• Experience with PDF and Image processing libraries (e.g. PyMuPDF, OpenCV, Pillow)
• Experience in Machine Learning/Data Science (ML algorithm selection, feature engineering, model training, hyperparameter tuning, supervised and unsupervised learning, building model pipelines, using ML tools/libraries/frameworks)
• Experience working with AWS big data technologies (Redshift, S3, EMR, Glue)

• Design and implement end-to-end document intelligence pipelines on AWS
• Develop and optimize ML models for document classification, segmentation, and field extraction
• Build scalable data processing systems handling PDFs up to 2000 pages
• Collaborate with subject matter experts to create and refine requirements for extraction
• Own features from research through production deployment and monitoring
• Establish evaluation frameworks and quality metrics for extraction accuracy