Job Description

Key Responsibilities

• Design and develop Reinforcement Learning models to optimize collections strategies, customer treatment paths, and recovery outcomes.
• Build adaptive decisioning systems using techniques such as:
  • Q-Learning
  • Deep Q Networks (DQN)
  • Policy Gradient Methods
  • Contextual Bandits
  • Markov Decision Processes (MDP)
• Develop sequential and behavioral models for customer engagement, repayment prediction, and collections prioritization.
• Apply stochastic modeling and probabilistic methods to optimize dynamic treatment strategies under uncertainty.
• Collaborate with business stakeholders to translate collections and risk management problems into scalable AI/ML solutions.
• Build and maintain machine learning pipelines in Databricks or similar distributed computing environments.
• Conduct experimentation, simulation, and offline policy evaluation to validate RL strategies before deployment.
• Work with large-scale structured and unstructured datasets to derive actionable insights and improve operational performance.
• Partner with engineering and MLOps teams to deploy and monitor production-grade ML/RL models.
• Mentor junior data scientists and promote best practices in modeling, experimentation, and AI governance.

Must-Have Qualifications

• Strong experience in Reinforcement Learning and sequential decision-making systems.
• Hands-on expertise with:
  • Reinforcement Learning algorithms (Q-Learning, DQN, PPO, Bandits, etc.)
  • Markov Decision Processes (MDP)
  • Stochastic modeling and probabilistic systems
  • Machine learning and predictive modeling
  • Experimentation and simulation frameworks
• Strong programming skills in Python and SQL.
• Experience with Databricks, Spark, or similar big data/cloud analytics platforms.
• Experience building scalable ML pipelines and deploying models into production environments.
• Strong understanding of feature engineering, model validation, and performance optimization.
• Ability to communicate complex AI/ML concepts to technical and non-technical stakeholders.

Preferred / Good-to-Have Skill

• Experience in collections, credit risk, customer analytics, or financial services domains.
• Familiarity with:
  • Deep Learning frameworks (TensorFlow, PyTorch)
  • MLOps and CI/CD workflows
  • Real-time decision systems
  • Cloud platforms such as AWS, Azure, or GCP
• Exposure to causal inference, uplift modeling, or optimization techniques.
• Knowledge of customer lifecycle analytics and behavioral segmentation.
• Experience working in Agile delivery environments.

• Strong experience in Reinforcement Learning and sequential decision-making systems.
• Hands-on expertise with:
  • Reinforcement Learning algorithms (Q-Learning, DQN, PPO, Bandits, etc.)
  • Markov Decision Processes (MDP)
  • Stochastic modeling and probabilistic systems
  • Machine learning and predictive modeling
  • Experimentation and simulation frameworks