About the Role We build machine learning models end-to-end.You will take full ownership of one of our core revenue-generating risk and underwriting models.This will involve any and all of the following workstreams: Data collection Data pipelines Retraining and optimization Evaluation Inference and serving Monitoring and quality assurance Other related dev tooling and infrastructure work You will conduct ad-hoc analyses, tests and retrospectives on questions related to our models, for delivery to internal stakeholders and external partners/clients You will serve as a go-to expert in title risk and underwriting for our team and the company. This includes exploring ways to reduce model losses, improve the fidelity of our projected losses, and identify new risks in our domain You will deliver new data science products targeting unexplored areas of the title and escrow workflow, from their inception to their deployment as machine learning models Required Qualifications 5+ years of professional experience in data science or machine learning engineering Strong background in statistics, computer science, machine learning, operations research, or mathematics Experience developing and deploying models end to end in a production environment, with proficiency in Python, SQL, Git, Docker Experience with object-oriented programming and software engineering practices, such as version control, unit and integration testing, CI/CD and API development Enthusiasm about taking ownership of complex projects, including managing the requests coming from multiple stakeholders throughout the company Ability to manage stakeholders and communicate technical concepts to people throughout the company, including business partners in product and senior leadership. Nice-to-Have Qualifications Experience in the domains of risk, credit, or insurance PyTorch, Transformers, LLMs (GPT, Claude, Gemini, Llama, Mistral, etc) Airflow, Kubernetes and MLOps Business intelligence tools such as Looker Cloud platforms such as Azure, AWS, etc
- li-remote
- J-18808-Ljbffr
integration-testing statistics Amazon Web Services (AWS) operations-research Computer Science Azure data-collections OpenAI GPT Communication Looker MLOps Docker Retrospective meetings data-pipelines Airflow Unit testing version-control Machine Learning evaluation Google Gemini stakeholder-management OOP Git CI/CD mathematics Python testing api-design Software Development Engineer risk management SQL Kubernetes PyTorch LLM