We are looking for a ML / MLOps Engineer to contribute to production-grade AI systems within a fast-paced technology organization (insurance technology company). The team owns multiple customer-facing and internal AI applications, including real-time decisioning, operational automation, chatbots, and ML infrastructure. This is a production ML engineering role. Candidates who primarily focus on offline modeling or hand models off to other teams for deployment are not a good fit.

Responsibilities

• Design and build APIs and pub/sub event streams to support real-time machine learning inference and automated agentic processes.
• Play a role in the development and maintenance of both online and offline feature stores for machine learning.
• Gain familiarity with the property casualty insurance sector, including key policyholder and product attributes, to help enhance model effectiveness.
• Implement industry-standard MLOps and LLMOps techniques to monitor ML models, feature sets, and agentic systems for performance degradation and data drift.
• Support the ongoing development of our core MLOps platform, as well as the codebase and infrastructure for serverless AI applications.
• Validate the performance of machine learning models through rigorous training and testing methodologies.
• Collaborate with Data Science teams to engineer new features, construct transformation pipelines, integrate custom loss functions, and experiment with novel inference strategies such as chaining and shadow deployments.
• Create and scale new agentic AI automations, guiding them from initial proof-of-concept through to full production deployment.
• Construct evaluation frameworks designed to rigorously test AI applications, covering not only standard workflows but also the complex, real-world scenarios common to the car insurance domain.
• Utilize the Python data ecosystem to execute machine learning projects and initiatives.
• Take part in the team's weekly on-call rotation, addressing alerts promptly to maintain high service availability for both customers and internal stakeholders.

Requirements

• Experience with Python (production-quality code)
• Experience with Python data science and machine learning libraries, including scikit-learn, pandas, numpy and related libraries.
• Hands-on experience deploying and operating ML models in production.
• Hands-on AWS experience (Lambda, Step Functions, DynamoDB, IAM, containers)
• Experience with Kafka or other event-driven systems
• Experience deploying ML models to production
• Git / CI/CD experience

Nice to Have

• Experience with MLOps platforms and automation tools
• Real-time data pipelines
• Experience with AI chatbots or retrieval-augmented generation (RAG) systems