This role pays less than
94%
of similar roles. Most pay
$160,000–$242,112
— the shaded band above.
At the midpoint, this role pays about
$127k
versus about
$201k
for comparable roles.
Based on 240 similar postings.
Employer
About Humana
Humana Inc. is a leading American for-profit health insurance company based in Louisville, Kentucky, specializing in Medicare Advantage plans.
Humana currently has
20 open roles
on FindRole.
Listed pay typically runs
$129,300–$177,800
across 20 roles with salary data.
The Senior Decision Intelligence Engineer joins the team to design, train, and improve the reinforcement learning policy powering the Next Best Action platform. This hands-on, research-oriented role involves designing and evaluating decision-making algorithms while instrumenting training pipelines in collaboration with data and platform engineers. The engineer ensures systems operate within clinical eligibility rules and specific program objectives. Key responsibilities include managing full software development cycles including front-end, back-end, and database integrations. Technical requirements include proficiency in Python 3.x, PyTorch or TensorFlow, Ray RLlib, Databricks, PySpark, Delta Lake, and MLflow. The role requires expertise in reinforcement learning, operations research methods like linear programming, and simulation frameworks such as Gymnasium or SimPy. The work addresses the challenge of creating reliable, auditable decision systems within regulated healthcare environments where member safety and explainability are critical requirements.
Design, train, and improve the reinforcement learning policy for the Next Best Action platform.
Develop and evaluate decision-making algorithms to optimize automated outcomes.
Instrument and manage large-scale training pipelines using tools like Ray RLlib and MLflow.
Ensure systems operate within clinical eligibility rules and specific program objectives.
Manage full-stack development tasks including front-end, back-end, and database integrations.
Diagnose failure modes in learned policies such as instability and distributional shift.
Build simulation environments for policy evaluation and backtesting using frameworks like Gymnasium or SimPy.
Maintain production-ready systems that ensure auditability and explainability in a regulated healthcare environment.
What we're looking for
Must have 5+ years of experience building and operating large-scale production systems like recommendation engines or optimization platforms.
Must have 2+ years of experience implementing reinforcement learning, operations research methods, or simulation-driven decision systems in production.
Must possess deep familiarity with Markov Decision Processes, Bellman equations, reward shaping, and constraint formulation.
Must be proficient in Python 3.x and experienced with PyTorch or TensorFlow for model implementation.
Must have experience using Ray RLlib or equivalent distributed frameworks for large-scale training.
Must have experience with Databricks, PySpark, and Delta Lake for processing large-scale data pipelines.
Must have experience with MLflow for experiment tracking and model registry management.
Must be able to work from a dedicated space that protects member PHI/HIPAA information.
Databricks
Spark
Delta Lake
Python
SQL
Medallion Architecture
Kafka
Azure Data Factory
Azure Data Lake Storage
Azure Event Hubs
Unity Catalog
Delta Live Tables
Great Expectations
Monte Carlo
Feature Engineering
Data Modeling