Utkarsh Sinha

About Utkarsh Sinha

Utkarsh Sinha is a Senior Research Engineer leading advanced R&D at Xecta Digital Labs, where he develops hybrid physics-informed machine-learning (PIML) models for reservoir simulation, forecasting, and field-scale optimization across unconventional and conventional assets. His work integrates multiphase-flow physics, operational constraints, and advanced ML architectures to deliver robust, scalable digital solutions deployed across thousands of wells. In parallel, Utkarsh serves as a long-standing Volunteer Research Associate with Dr. Birol Dindoruk at the IPB & F Consortium (Texas A&M University). His research spans PVT/EOS modeling, miscible gas injection, phase behavior, and physics-guided surrogate models, including hybrid MMP correlations, viscosity modeling, and web-based PIML applications that combine domain fundamentals with data-driven inference. Utkarsh holds a Master’s in Petroleum Engineering from the University of Houston (2017–2018), supported by the BP Student Scholarship and the Engineering Dean’s Master Scholarship in recognition of outstanding academic performance.🤝 He previously served as Board Advisor to the SPE Data Science and Engineering Analytics (DSEA) Technical Director (2020–2024), guiding strategy, content development, and cross-disciplinary integration. He also serves on the PetroWiki Steering Committee, supporting taxonomy expansion, content quality, and inclusion of emerging digital-workflow domains. Utkarsh has been a Technical Committee Member and Session Chair for the SPE Improved Oil Recovery Conference (2022, 2024, 2026) and has held multiple roles across URTeC (2022–2026), including Reviewer, Session Chair, and Theme Chair. He most recently led the “Applications of AI & Data Science” theme for URTeC 2026 and chaired the PIML Special Session, shaping review frameworks, session structure, and emerging-topic prioritization across digital subsurface workflows. He is an active reviewer for SPE Journal, SPE Production & Operations, Geoenergy Science and Engineering, Journal of Chemical Information and Modeling, Upstream Oil & Gas Technology, and Environmental Technology & Innovation. His contributions to data-science applications in petroleum engineering include publications in leading journals (SPEJ, JPSE, Geoenergy Science & Engineering), major conference proceedings (ATCE, IOR, URTeC, ADIPEC), and six patents (one granted, US Patent 11,767,750). With 240+ citations, his work reflects deep technical impact and sustained innovation within the subsurface digital-engineering domain.

Experience

1. Senior Research Engineer

   Xecta

   Aug 2023 — Present · Houston, TX, US

   Lead R&D initiatives to develop physics-informed-machine-learning (PIML) models for reservoir simulation and production optimization, enabling digital transformation across conventional and unconventional assets. Authored 15 conference papers and contributed to 6 patents as co-inventor.• Develop hybrid frameworks integrating reservoir and production engineering with advanced data-driven methods and operational constraints for field-level decision support. Built a multiphase rate-forecasting model trained on 10 years of data from 100 wells in under an hour, enabling rapid, high-fidelity predictions for shale development planning. Integrated an inter-well interference detection and quantification workflow with production forecasting, supporting a $1.5M/year commercial solution deployed across shale wells. Built an automated liquid-loading diagnosis and mitigation workflow across gas wells, yielding 5–7% incremental production uplift per well and driving broad commercial adoption.• Deploy advanced data-driven techniques—including PINNs, SINDy, autoencoders, and dynamic mode decomposition—tightly coupled with physics principles and field-level operational constraints.• Execute reservoir modeling and production optimization workflows for integrated asset management in offshore conventional reservoirs. Principal modeler for Xecta’s hybrid reservoir simulator (RGNet), integrating physics-based and data-driven methods to advance next-generation simulation capabilities and deliver significant client impact.• Enhance model performance and scalability using vectorization, parallelization, and optimized data pipelines.• Lead cross-functional teams on R&D POCs, driving model development, validation, and deployment from pilot to full-scale implementation with measurable field impact. Technical lead for an offshore water-production forecasting POC in the Middle East, transitioning the pilot to a $700K/year, 5-year deployed product;

Education

• Vellore Institute of Technology

  Bachelor of Technology (BTech), Chemical Engineering

  2012 — 2016

• University of Houston

  M.Eng., Petroleum Engineering

  2017 — 2018