Data Scientist

Hero Fincorp; Gurugram District; On-site; INR 1,800,000 - 3,000,000; Full time

Job summary

Hero Fincorp in India is seeking a data science professional focused on generative AI and hyper-personalization to build risk and customer engagement models.

You will design scalable feature pipelines, deploy models in production with Docker/Kubernetes, and collaborate with risk, product, and data teams to translate business needs into scalable AI solutions.

Qualifications

• 4+ years of hands-on experience in data science with a focus on generative AI and hyper-personalization.
• Bachelor's or master's in computer science, engineering, data science, ML, statistics or related field.
• Proficiency in programming languages like Python/R/Java/PySpark.
• Experience with deep learning frameworks such as TensorFlow, PyTorch, or Keras.

Responsibilities

• Develop and maintain data science models for credit risk, customer behavior, income estimation, and surrogate modelling.
• Build scalable feature pipelines using Python/PySpark/SQL with large datasets.
• Deploy models into production systems using Docker/Kubernetes and integrate with decision APIs.
• Implement MLOps practices: testing, monitoring, drift detection, versioning, and lifecycle management.
• Work with AWS-based or equivalent cloud environments to optimize pipelines and costs.
• Collaborate with risk, product, and data engineering teams to translate requirements into scalable AI solutions.

Skills

Python

Pyspark

R

Java

TensorFlow

PyTorch

Keras

Pandas

NumPy

Scikit-learn

polars

AWS

GCP

Azure

Data visualization

Education

Bachelor's/Master's in CS/Engineering/Data Science/ML/Statistics

Job description

Key Responsibilities:

• Credit Risk & Decisioning Models
  • Develop and maintain models for:
  • Credit risk (PD, delinquency, segmentation)
  • Customer behaviour (collections, engagement, response)
  • Income estimation / surrogate modelling (where applicable)
  • Work with bureau data (e.g., CIBIL/CRIF), transactional data, and alternative data sources
  • Translate business policies into model-driven decision frameworks.
• Data Engineering & Feature Pipelines
  • Build scalable feature pipelines using:
  • Python / PySpark / SQL
  • Handle large-scale datasets (structured + semi-structured)
  • Implement robust feature engineering (e.g., bureau features, exposure, EMI, leverage, etc.)
• Model Deployment & Integration
  • Deploy models into production systems using:
  • Docker, Kubernetes, or similar containerization tools
  • Integrate models with decision engines / APIs
  • Ensure low-latency and high-throughput inference pipelines
• MLOps, Testing & Monitoring
  • Implement:
  • Integration testing for model pipelines
  • Stress testing (performance under load)
  • Model monitoring (drift, stability, performance)
  • Work with DevOps teams to ensure reliable production systems
  • Handle versioning, rollback strategies, and model lifecycle management.
• Cloud & Infrastructure
  • Work on AWS-based environments (or equivalent cloud platforms)
  • Understand distributed systems and compute optimization
  • Optimize pipelines for performance and cost
• Cross-functional Collaboration
  • Work closely with:
  • Risk teams
  • Product / business stakeholders
  • Data engineering & platform teams
  • Translate business requirements into scalable technical solutions

Eligibility Criteria for the Job

Experience

4+ years of hands-on experience in data science, with a focus on generative AI and hyper-personalization

Education

Bachelors/masters in computer science, Engineering, Data Science, ML, Statistics, or related field

Skills & Competencies

• Proficiency in programming languages like Pyspark, Python, R, or Java.
• Experience with deep learning frameworks such as TensorFlow, PyTorch, or Keras.
• Strong knowledge of data processing libraries (e.g., Pandas, NumPy, Scikit-learn, polars).
• Familiarity with cloud platforms (AWS, GCP, Azure) for deploying models and managing data pipelines.
• Experience with data visualization tools (e.g., Matplotlib, Seaborn, Tableau).