About The Team
We are looking for a Deep Learning Engineer to drive the evolution of our risk modeling capabilities. This role goes beyond traditional ML — you will be at the forefront of transitioning production risk systems from classical approaches to modern deep learning architectures, building large-scale pre-trained models, and applying reinforcement learning for decision making.
You will work closely with Risk, Data Science, and Engineering teams to design and deliver next-generation models that power credit decisions at scale.
Job Description
- Design and implement deep learning models to enhance existing tree-based risk modeling pipelines, with a focus on improved generalization, representation learning, and scalability.
- Develop and optimize large-scale pre-trained models, including architecture design, pre-training strategies, fine-tuning, and inference optimization for production environments.
- Build reinforcement learning systems for dynamic credit limit and interest rate adjustment, including reward shaping, policy optimization, and online learning frameworks.
- Engineer deep learning embedding solutions for heterogeneous data sources (e.g., bureau data), extracting rich latent representations to improve downstream model performance.
- Develop sensitivity modeling for in-loan pricing decisions, capturing complex user-level behavioral responses to rate and limit changes.
- Collaborate with Risk and Data Science teams to translate business problems into DL problem formulations, evaluating modeling trade-offs across accuracy, latency, and fairness.
- Build and maintain end-to-end model pipelines covering training, evaluation, deployment, and monitoring across batch and real-time systems.
- Partner with Data Engineering to ensure feature consistency, data quality, and reliable offline-to-online feature parity.
- Contribute to modeling best practices, reproducibility standards, and internal technical documentation.
Requirements
- Master's degree in Computer Science, Mathematics, Statistics, or a related quantitative field.
- 3+ years of hands-on experience in applied deep learning or machine learning engineering, with production deployment experience.
- Proficiency in Python and deep learning frameworks (PyTorch preferred; TensorFlow/JAX a plus).
- Strong understanding of neural network architectures including Transformers, sequential models, and embedding-based models.
- Practical experience with at least one of: reinforcement learning (policy gradient, actor-critic), large-scale pre-training / fine-tuning, or representation learning.
- Experience in model monitoring, drift detection, and lifecycle management in production.
- Experience in fintech, credit risk, or financial services is strongly preferred.
- Strong communication skills to collaborate with cross-functional stakeholders across Risk, Product, and Engineering.
Nice to Have
- Experience with large-scale distributed training (e.g., multi-GPU, parameter servers).
- Familiarity with online learning or continual learning systems.
- Exposure to causal inference or uplift modeling for pricing/limit optimization.
- Experience with feature stores or real-time serving infrastructure.