Lead Decision Intelligence Engineer - NBA

Location: Remote nationwide in the United States; hybrid work is possible, preferably based in Boston, MA. Occasional travel to hirify.global offices for training or meetings may be required.

Salary: $129,300–$177,800 per year, plus eligibility for a bonus incentive plan.

Company

hirify.global is a U.S. healthcare and insurance company serving members through hirify.global insurance services and CenterWell healthcare services.

What you will do

• Design, implement, and evaluate reinforcement learning algorithms for the Next Best Action platform, including policy-gradient, value-based, and offline RL methods.
• Define member state representations, action spaces, reward structures, and constraint mappings for clinical eligibility, suppression rules, and program objectives.
• Build simulation and backtesting environments and diagnose policy collapse, credit assignment issues, and distributional shift before production deployment.
• Own nightly Databricks training workflows, including PySpark feature engineering, distributed Ray RLlib training, batch scoring, and reproducible data pipelines.
• Manage MLflow experiment tracking, model registry, artifacts, versioning, evaluation gates, and rollback capability.
• Integrate policy outputs with decision engines, Rules Engine services, Redis, Kafka feedback loops, and clinical and compliance governance processes.

Requirements

• Bachelor's degree in computer science or a related field.
• 8+ years of software engineering experience building large-scale production systems, plus 3+ years implementing reinforcement learning or deep learning systems in production.
• Strong knowledge of the Bellman equation, reward shaping, exploration-exploitation tradeoffs, constraint mapping, and RL-specific failure modes.
• Proficiency in Python and experience with PyTorch or TensorFlow, Ray RLlib, Databricks, PySpark, Delta Lake, and MLflow.
• Experience shipping reliable ML systems operating at production scale and processing data for millions of users.
• Work visa sponsorship is not available; candidates must be able to work in the United States.

Nice to have

• Experience with multi-agent reinforcement learning, PettingZoo, Gymnasium, probabilistic modeling, Markov Decision Processes, or linear programming.
• Experience operating RL systems in healthcare, finance, insurance, or other regulated domains.
• Familiarity with Kafka feedback loops and OpenTelemetry instrumentation for ML training observability.

Culture & Benefits

• Remote or hybrid work with typical Monday–Friday business hours and a 40-hour workweek.
• Dedicated home workspace and internet service with at least 25 Mbps download and 10 Mbps upload required for home or hybrid work.
• Medical, dental, and vision benefits, 401(k), paid time off, holidays, parental and caregiver leave, disability coverage, and life insurance.
• Work must protect member PHI and HIPAA information through a secure, interruption-free workspace.

Hiring process

• Initial application prescreen followed by an on-demand HireVue assessment with predetermined questions lasting approximately 10–15 minutes.
• Applicants who progress after the assessment are invited to subsequent interview rounds.
• SSN entry into Workday may be requested after a formal employment offer for duplicate-profile screening.