Lead ML/AI Engineer

Title: Lead ML/AI Engineer Location: United States

About the Role

TailorCare's Data Science team builds the ML/AI systems that decide which patients we reach, when, and how. We are shipping a portfolio of models and a real-time decision platform that puts them to work across every market we serve and learns from every outcome it touches. Model outputs become events, events reach care teams and payer systems, outcomes come back days later, and every piece has to be versioned, replayable, monitored, and affordable under PHI controls. This role owns the engineering that makes that work in production.

As Lead ML/AI Engineer, you will be the technical lead for the platform that turns model output into action. You will set the engineering standards for our decisioning, streaming, and ML release systems, stay hands-on in the highest-risk components, and hold releases to a production bar through design and code review. You will own how our Databricks lakehouse and AWS-native services fit together as the platform evolves. We are looking for someone who has built systems that decide actions and prove they paid off, not just models that predict. This is a senior individual contributor role. Reporting to the Director, AI & Decision Intelligence, you will work closely with data science, data engineering, software engineering, medical economics, clinical, and operations partners.

Primary Responsibilities

• Own the decisioning and feedback layer between score and action: action selection and ranking, holdout and experiment assignment, outcome capture, the production ROI calculation, and the pipelines that measure lift per action and channel.
• Own the engineering implementation, auditability, and reproducibility of the production savings calculation, in partnership with medical economics and data science, who own the methodology.
• Set the design for event-driven, real-time integrations between model output and the systems that act on it: streaming pipelines, versioned schemas and contracts, replay, ordering, idempotency, and failure handling. Own the contracts with partner teams.
• Set the cost-to-serve standard for the platform: define per-model and per-decision budgets for compute, streaming, inference, and LLM spend, and  make cost a release criterion alongside latency and accuracy.
• Define the production bar for ML models and hold releases to it: CI/CD, model registry, alerting, drift, coverage, and calibration by cohort, market, and payer. Engineer the promotion gate a model or ranking must pass before it goes live.
• Own the target architecture and technical roadmap for our AWS-native footprint alongside our Databricks lakehouse, including  infrastructure as code, IAM and secure data handling, containerized inference, and managed foundation-model services under PHI controls.
• Set technical direction and production standards for our production GenAI applications, including LLM pipelines and agentic workflows, with the evaluation, retrieval, guardrail, and cost controls needed to run them safely on patient data.
• Partner with data scientists on modeling and measurement methodology, with medical economics on how savings is defined, and with data and software engineering on shared pipelines, interfaces, and contracts.
• Own engineering standards for the team: code review, automated testing, reproducibility, documentation, and release discipline. Lead design reviews, document architecture decisions, and mentor engineers and data scientists on system design and production practice.
• Represent the team's technical position with R&D leadership, finance, and clinical and operations partners.
• Other duties as assigned

Qualifications

Required

• Bachelor's degree in Computer Science, Engineering, or a related field, or equivalent experience.
• 8+ years in ML engineering, MLOps, or software engineering, including 4+ years operating ML systems in production and 2+ years as the technical lead or architect of a production ML or decisioning system.
• Demonstrated experience building or operating prescriptive systems in production: recommendation, decisioning, prioritization, or experimentation platforms that allocate a constrained resource against a measured outcome.
• Production experience with AWS data and ML services and infrastructure as code (Terraform, CDK, or CloudFormation).
• Production experience with Databricks, PySpark, Delta Lake, MLflow, or equivalent.
• Hands-on experience designing and running event-driven or streaming systems (Kinesis, Kafka, or equivalent) with schema or contract management.
• Strong Python and SQL, and hands-on experience running ML workloads in production: training pipelines, model serving, feature systems, or ML platform tooling.
• CI/CD for ML workloads and model and pipeline monitoring, including calibration across heterogeneous populations.
• Track record of setting engineering standards and raising the bar on a team through review and mentorship.
• Ability and willingness to travel up to 10% as needed for onsite meetings, team collaboration, and company events.
• Ability and willingness to travel up to 10% as needed for onsite meetings, team collaboration, and company events.

Preferred

• Building and evaluating ML models with scikit-learn, XGBoost, or PyTorch.
• Healthcare economics or value-based care exposure: savings methodology, attribution, payer contract constraints.
• dbt and lakehouse data modeling.
• Healthcare data (claims, pharmacy, eligibility, prior authorization, EHR) and PHI handling requirements.
• Uplift modeling, bandit or sequential decisioning, or holdout-based measurement in production.
• Containerized model serving and staged or canary deployment patterns.
• Amazon Bedrock or other managed foundation-model platforms under a HIPAA BAA.
• Agentic frameworks and retrieval systems in production.
• Experience building or operating LLM-based applications, including evaluation and guardrails.

Skills 

• Engineering Mentality – Finds the common denominator across work and builds the platform that makes the next model faster and safer to ship than the last.
• Scientific Approach – Treats a design as a hypothesis, instruments it, and changes course when the data says so.
• Seek Simplicity – Builds with the minimum necessary complexity and designs for the person who maintains it next.
• Ownership – Takes a system from design to production to monitoring and answers for whether it works in the market, not just whether it shipped.
• Collaborative – Works across data science, data engineering, software engineering, medical economics, clinical, and operations to agree on interfaces and hold to them.
• Good Communication – Explains what a system does, what it costs, and why it matters in terms a care team or finance partner can act on.