Group Data Scientist

Ehpv; Bengaluru; On-site; INR 1,800,000 - 3,600,000; Full time

Job summary

Ehpv specializes in designing ML and decision-science solutions for optimization problems in Bengaluru. The role involves translating real-world challenges into formal mathematical formulations and rapidly prototyping using agentic coding tools, while upholding strong engineering discipline.

You'll lead model selection, experimentation strategy, and collaboration with ML engineers, with responsibilities spanning productionization, monitoring, and clear communication of outcomes to stakeholders.

Qualifications

• Experience applying ML to decision-making or optimization.
• Proficiency with agentic coding assistants and AI-supported workflows.
• Advanced Python and ML frameworks experience.
• Strong ability to structure messy business problems into tractable solutions.
• Excellent communication with technical and business stakeholders.
• Expertise in Python, PyTorch, OR-Tools, RL libraries, SQL, Docker, Git, MLflow and cloud platforms.

Responsibilities

• Lead design and development of ML and decision-science solutions for optimization.
• Translate real-world challenges into well-defined mathematical formulations.
• Prototype rapidly using AI-assisted tools and modern workflows.
• Productionize models with attention to latency, monitoring, and safe deployment.
• Document methodologies and communicate recommendations to stakeholders.
• Provide technical leadership and mentor team members.

Skills

Machine Learning

Optimization

Python

PyTorch

TensorFlow

Statistics

Experiment design

Communication

AI-assisted development

Tools

OR-Tools

RL libraries (Ray RLlib, Stable Baselines)

Docker

Git

MLflow

Job description

KEY ACCOUNTABILITIES

• Lead the design and development of ML and decision-science solutions for high-impact operational problems, including planning, sequencing, routing, allocation, and resource optimization.
• Translate ambiguous real-world challenges into well-defined mathematical, algorithmic, or learning formulations with clear objectives, constraints, and measurable success metrics.
• Rapidly prototype and iterate using agentic coding tools and modern development workflows to accelerate experimentation, code generation, refactoring, and test creation while preserving strong engineering discipline.
• Develop, benchmark, and improve models across areas such as:
  • Optimization and solver-based methods:MILP, CP-SAT, constraint programming, heuristics, metaheuristics, and search-based techniques
  • Decision Intelligence and Reinforcement Learning:contextual bandits, offline RL, deep RL, Monte Carlo Tree Search, policy learning, and value-based methods
  • Predictive ML:forecasting, estimation, and probabilistic models that support downstream decision systems
• Design rigorous evaluation frameworks, including simulation environments, counterfactual analysis, ablation studies, stress testing, and scenario-based performance assessment.
• Define KPIs, acceptance criteria, and experimentation standards to ensure solutions are both scientifically sound and operationally relevant.
• Partner closely with ML engineers and platform teams to productionize models, with attention to latency, throughput, reproducibility, monitoring, versioning, and safe deployment practices.
• Provide technical leadership in model selection, experimentation strategy, and research direction, while mentoring less experienced scientists and raising the quality bar across the team.
• Document methodologies, assumptions, results, and trade-offs clearly, and communicate recommendations effectively to both technical and business stakeholders.
• Strong experience applying machine learning and algorithmic methods to real-world decision-making or optimization problems.
• Demonstrated proficiency with agentic coding assistants and AI-supported development workflows to accelerate research and engineering output without compromising code quality, maintainability, or testing standards.
• Advanced Python skills and strong hands-on experience with ML frameworks such as PyTorchpreferred, or TensorFlow.
• Solid grounding in algorithms, optimization, probability, statistics, and experimental design.
• Proven ability to structure messy, high-ambiguity business problems into tractable technical solutions with measurable impact.

• Strong communication skills, with the ability to explain complex technical concepts, experimental findings, and trade-offs to diverse stakeholders.

QUALIFICATIONS, EXPERIENCE AND SKILLS

• Strong experience applying machine learning and algorithmic methods to real-world decision-making or optimization problems.
• Demonstrated proficiency with agentic coding assistants and AI-supported development workflows to accelerate research and engineering output without compromising code quality, maintainability, or testing standards.
• Advanced Python skills and strong hands-on experience with ML frameworks such as PyTorchpreferred, or TensorFlow.
• Solid grounding in algorithms, optimization, probability, statistics, and experimental design.
• Proven ability to structure messy, high-ambiguity business problems into tractable technical solutions with measurable impact.
• Strong communication skills, with the ability to explain complex technical concepts, experimental findings, and trade-offs to diverse stakeholders.

• Expertise in Python, PyTorch, OR-Tools and solver stacks, RL libraries such as Ray RLlib or Stable Baselines, SQL, Docker, Git, MLflow, and cloud platforms.