BC Sector O and G-Data Scientist-Senior Manager
EY · Bengaluru
At EY, you’ll have the chance to build a career as unique as you are, with the global scale, support, inclusive culture and technology to become the best version of you. And we’re counting on your unique voice and perspective to help EY become even better, too. Join us and build an exceptional experience for yourself, and a better working world for all.
EY GDS | PRINCIPAL DATA SCIENTIST
Oil & Gas | Senior Manager
Job Title Principal Data Scientist – Senior Manager
Experience Level 15+ Years
Job Description
At EY, you’ll have the chance to build a career as unique as you are, with the global scale, support, inclusive culture and technology to become the best version of you. We’re counting on your unique voice and perspective to help EY become even better. Join us and help build a better working world.
We are seeking a hands-on Principal Data Scientist to solve complex modelling problems across upstream and downstream Oil & Gas. You will formulate operational challenges, establish defensible scientific approaches, develop critical modelling components, and define the evidence required for reliable deployment across wells, fields, equipment, process units, and sites.
Role Details
Key responsibilities include:
Scientific formulation and advanced modelling
- Formulate operational problems as predictive, diagnostic, causal, or prescriptive tasks, defining targets, objective functions, constraints, decision variables, acceptance criteria, and evaluation methods.
- Assess data sufficiency, measurement quality, sampling bias, confounding, identifiability, and operating coverage to determine whether available evidence supports the intended outcome.
- Select statistical, machine learning, simulation, or hybrid approaches suited to noisy measurements, heterogeneous assets, sparse labels, rare events, censored histories, and distribution shifts.
- Apply probabilistic or hierarchical modelling, uncertainty quantification, survival analysis, semi-supervised learning, transfer learning, or domain adaptation where appropriate.
- Develop methods for multivariate forecasting, state estimation, change detection, degradation modelling, and industrial time-series analysis while distinguishing process dynamics from measurement artefacts.
- Combine learned models with engineering equations, mass and energy balances, simulation outputs, or physical constraints, evaluating accuracy, plausibility, extrapolation behaviour, and uncertainty propagation.
- Connect predictions with constrained optimization, assessing feasibility, sensitivity, robustness, and decision quality under uncertainty; design experiments or causal analyses when observational evidence is insufficient.
Upstream applications
- Define modelling approaches for production forecasting, well surveillance, artificial-lift performance, drilling analytics, and production-system optimization.
- Separate the effects of reservoir behaviour, operating constraints, equipment condition, interventions, formation changes, and operating practices when interpreting performance.
- Collaborate with reservoir engineers and geoscientists on property prediction, uncertainty analysis, and simulation-assisted modelling.
- Determine whether models can transfer between wells or fields and define the data, recalibration, adaptation, and validation required.
Downstream applications
- Define modelling approaches for product-quality prediction, yield modelling, equipment degradation, blending, and process or energy optimization.
- Account for feedstock composition, throughput, catalyst age, equipment condition, process configuration, and control-system behaviour.
- Combine data-driven methods with engineering relationships, process simulations, and mass and energy balances.
- Connect predictive models with operating, planning, blending, or optimization workflows within defined physical and operational constraints, validating across feedstock changes, campaigns, grades, and configurations.
Model assurance and generalization
- Establish standards for baselining, benchmarking, feature-ablation studies, calibration, sensitivity analysis, robustness, and out-of-distribution evaluation.
- Challenge apparent improvements caused by leakage, weak baselines, inappropriate metrics, selective testing, or inadequate engineering interpretation.
- Distinguish models that reproduce historical associations from models capable of evaluating proposed operating changes.
- Define applicability limits, uncertainty treatment, monitoring requirements, fallback behaviour, and recalibration or retraining criteria.
- Assess predictive performance separately from recommendation feasibility and quality, and establish adaptation and validation strategies across assets, sites, configurations, and operating regimes.
Technical leadership and scalable delivery
- Remain hands-on in experiments, model development, prototyping, and troubleshooting.
- Resolve failures across data preparation, measurement quality, engineering assumptions, model formulation, software implementation, and production behaviour.
- Develop reusable modelling components, adaptation strategies, technical standards, and reference implementations; design reproducible experiments whose conclusions can be reviewed and defended.
- Evaluate emerging techniques through controlled comparisons with established methods and review production implementations for consistency with the intended scientific approach.
- Conduct scientific reviews, challenge modelling decisions, mentor Lead Data Scientists, and collaborate with engineering, data, ML engineering, and product teams to integrate validated models into operational workflows.
Must Have Qualifications & Experience
- A degree in a relevant engineering, scientific, or quantitative discipline.
- Substantial hands-on experience developing and deploying Oil & Gas analytical solutions.
- Deep expertise in at least one upstream or downstream area, with working knowledge across both segments.
- Advanced expertise in at least one relevant applied modelling specialization.
- Strong foundations in statistics, machine learning, mathematical modelling, industrial time series, scientific computing, experimental design, and uncertainty assessment.
- Strong Python and SQL skills, including scientific prototyping, implementation of critical modelling components, and review of production code.
- Experience integrating engineering knowledge into model formulation and validation, together with reproducible experimentation, production ML, model generalization, and deployment across varied operating conditions.
- Evidence of personally resolving a difficult methodological problem and establishing an approach adopted by others.
- Experience conducting technical reviews and guiding multidisciplinary data science teams.
- Ability to defend technical decisions using quantitative evidence and clearly communicate assumptions, trade-offs, limitations, and operational implications.
Good to Have
- Delivery experience across both upstream and downstream Oil & Gas.
- A master’s degree, PhD, or relevant applied research experience.
- Expertise in system identification, robust optimization, physics-informed learning, simulation surrogates, or uncertainty propagation.
- Experience with distributed scientific computing, scalable inference, cloud ML platforms, and production MLOps.
- Experience combining validated numerical models with knowledge graphs, specialized language models, or agentic decision-support workflows.
- Publications, patents, open-source contributions, or reusable technical assets demonstrating applied scientific contribution.
What We Offer
- Opportunity to solve high-impact modelling and decision-support challenges across upstream and downstream Oil & Gas operations.
- A role with hands-on technical authority, scientific leadership, and the opportunity to establish reusable methods and standards for multidisciplinary teams.
- A collaborative environment spanning domain engineering, data science, ML engineering, product, and global EY teams.
- Access to EY’s global network, technology ecosystem, and learning opportunities while contributing to a better working world.
What We Look For
A hands-on technical authority who can explain a difficult Oil & Gas problem they personally solved, why established approaches were insufficient, how the alternative was formulated and validated, where it can and cannot generalize, and how the resulting approach enabled other data scientists to deliver reliable solutions.
EY | Building a better working world
EY exists to build a better working world, helping to create long-term value for clients, people and society and build trust in the capital markets.
Enabled by data and technology, diverse EY teams in over 150 countries provide trust through assurance and help clients grow, transform and operate.
Working across assurance, consulting, law, strategy, tax and transactions, EY teams ask better questions to find new answers for the complex issues facing our world today.