In the race to unlock AI’s full potential, data is both the fuel and the friction. As companies battle to integrate diverse control systems, legacy platforms, and siloed datasets, the question isn’t just how to centralize data, but how to make it usable, trustworthy, and AI-ready? With Generative AI offering new ways to structure and interpret messy data, is it time to rethink the traditional “clean first, deploy later” mindset? This session challenges conventional data strategies and explores how AI can reshape, not just rely on, your data foundations.

• How generative AI can help you move from data chaos to contextual clarity by structuring unstandardized data streams, reducing the burden of upfront data cleaning, and accelerating time-to-value in AI deployments
• How the “clean as you go” paradigm, based on iterative data refinement powered by real-time AI results, can uncover hidden issues, optimize resource allocation, and reshape how you engage vendors and suppliers
• The complexities of ownership, integration, and trust in data ecosystems, including challenges around data ownership, integration across control systems, and the pursuit of a single source of truth in a fragmented digital landscape 

Shivaprasad Sankesha Narayana

Lead Architect Consultant
bp

See Bio

Inderpreet Jalli

Senior Power Trading Analyst
NRG Energy

See Bio

Indu Nambiar

Principal, IT Architecture
California ISO

See Bio

In energy operations, critical knowledge is often buried in unstructured text, maintenance logs, field reports, technical manuals, and supplier documents. In this session, we will demonstrate how Natural Language Processing (NLP) can unlock that information and turn it into real-time operational insight. Discover how frontline teams are using NLP to reduce downtime, improve compliance, and make faster, more informed decisions, without needing to sift through thousands of documents. 

As digital transformation accelerates, the ability to scale AI across operations hinges on a solid data foundation with modernized data architecture. Yet many still struggle to move beyond siloed analytics and pilot projects due to fragmented systems, unclear ownership, and inconsistent data quality. This session explores how to identify high-impact pain points and design data architectures that support scalable, secure, and AI-ready analytics environments.

• Identifying operational pain points that are ripe for AI enablement, whether in maintenance, emissions tracking or production optimization
• Best practices for building a modern data architecture, including semantic layers, unified data pipelines, and governance frameworks that support cross-functional analytics
• Lessons learned on overcoming legacy system constraints and preparing data for AI 

As energy systems grow more complex and data-rich, the next frontier is not just prediction, but action. Agentic AI introduces autonomous, goal-driven agents capable of making decisions, coordinating tasks, and adapting to changing conditions in real time. This panel brings together operators who are exploring or deploying agentic AI to streamline operations, reduce manual intervention, and accelerate response times across asset-heavy environments.

• What agentic AI means for energy operations, from single-agent task automation to multi-agent orchestration across workflows
• How autonomous agents can optimize frontline decisions, such as dispatching maintenance crews, adjusting load forecasts, or managing emissions thresholds
• Implementation challenges including data integration, trust, oversight and aligning agent behavior with business goals
• Strategies for human-AI collaboration
• Lessons from early deployments, how agentic AI can reduce downtime, improve safety, and enhance operational agility 

David Moore

Global Wells Digital Advisor, Program Manager and AI Digital Delivery Lead
Shell

See Bio

As operations grow in complexity, the next frontier lies in orchestrating autonomous decision-making across entire business units. This session explores the theoretical and emerging practical applications of hierarchical multi-agent systems, a layered architecture of AI agents that collaborate, communicate, and self-organize to optimize workflows across drilling, subsurface, completions, and abandonment. 

As the industry seeks to move beyond reactive maintenance, machine learning offers a powerful opportunity to predict failures before they happen, reducing downtime, optimizing asset performance, and improving safety. This panel explores how early-stage pilots can lay the foundation for scalable, enterprise-wide predictive maintenance solutions.

• How to build a business case for machine learning in maintenance by identifying high-value assets, leveraging historical failure data, and demonstrating early wins
• Strategies for assessing data readiness and model adaptability, ensuring that predictive algorithms are trained on relevant, high-quality inputs and evolve with operational complexity
• Stakeholder engagement techniques that drive adoption, from localized decision-making to cross-functional collaboration between data teams and asset owners
• How proof-of-concept projects can be scaled across the enterprise with lessons learned in governance, infrastructure integration, and long-term value tracking 

Amit Kumar

Technical Project Manager, Data Science, Unconventional Business Line
ExxonMobil

See Bio

Sriram Ramaganesan

Director, APC, Blending and Optimization
Phillips 66

See Bio

Monitoring remote and offshore oil fields is a critical challenge. By replacing manual inspections and sensor-heavy setups with real-time video analytics, site surveillance can be digitized, improving uptime, reducing maintenance delays, and increasing productivity. The result: scalable, cost-effective field visibility without the need for constant human presence or complex sensor networks.  

As utilities modernize their infrastructure, the ability to extract actionable insights from visual and operational data is becoming a game-changer. In this session, Eversource Energy, New England’s largest utility, shares how they are deploying computer vision and machine learning to assist asset inspection, predictive maintenance, and internal engineering workflows for transmission and distribution systems.

Zefan Tang

Senior Data Scientist
Eversource Energy

See Bio

As operators face increasing pressure to improve safety and reduce downtime, AI-powered leak detection is emerging as a high-impact solution. This use case will illustrate how computer vision, thermal imaging, and edge computing are being combined to monitor pipelines in real time, enabling early identification of leaks without relying solely on traditional sensors.

Philippe Daroux

Operations Technology Product Line Manager
Chevron

See Bio

Timely, data-driven decisions are essential for improving reliability, reducing costs, and managing operational risk. This session focuses on how AI and advanced analytics are transforming asset management by enabling predictive maintenance and risk-based decision-making. Learn how to:

• Leverage machine learning models trained on historical and real-time data to detect early signs of equipment degradation, optimize maintenance intervals, and minimize unplanned downtime
• Apply risk-based frameworks to prioritize assets based on criticality and failure probability, ensuring that maintenance and capital investments are aligned with business impact
• Visualize AI-driven insights through integrated dashboards that combine IoT data, risk scores, and KPIs, empowering teams to act faster and with greater confidence 

As wells reach the end of their productive life, maximizing output becomes increasingly complex, especially for artificial lift systems like plunger lift and gas-assisted plunger wells. This session explores how machine learning is being used to optimize production by identifying the ideal shut in and flow times, gas injection rates, and equipment configurations. After a successful pilot, Ovintiv is moving toward full automation, allowing models to recommend optimal settings and prioritize wells for intervention.

Yousef Hamedi Shokrlu

Advisor, Data Science
Ovintiv

See Bio

In energy operations, visual data is everywhere, but often underutilized. Computer vision platforms powered by AI are changing that, enabling automated inspections, real-time risk detection and intelligent monitoring of assets. In this session, we’ll explore how computer vision is being used to enhance maintenance programs, improve the safety for frontline workers and deliver continuous visibility into operational environments. Discover how energy companies are leveraging these solutions to reduce manual effort, detect anomalies before they escalate, and drive smarter, safer decisions across the enterprise. 

As experienced engineers retire and operational demands grow, refineries are turning to AI to bridge the gap. In this session, Tim shares how anomaly detection is being used to surface early signs of equipment issues, enabling proactive maintenance, reducing unplanned downtime and supporting a shift towards predictive maintenance.

Using a dynamic risk analyzer, the team used a hands-off AI tool that automates data cleansing, modeling, and real-time monitoring. It flags subtle shifts in process behavior, such as changes in reactor temperature slopes, that may signal emerging equipment issues. By surfacing these early indicators, engineers can plan interventions more effectively and reduce the risk of costly failures.

The session also explores how to simplify model building and analysis, enabling engineers to extract insights from thousands of data points without writing code. While the pilot has shown promising results, human oversight remains critical, with a “trust but verify” approach guiding the path toward enterprise-wide adoption. 

Tim Sandford

Refining PI Coordinator
Marathon Petroleum

See Bio

As regulatory frameworks evolve and voluntary reporting commitments expand, organizations face mounting pressure to streamline how they collect, validate, and report operational data. This session explores how AI-powered compliance and reporting systems can dramatically reduce manual effort, enhance data integrity, and empower frontline teams to focus on high-value tasks. By embedding intelligence into reporting workflows, organizations can unlock new levels of agility, transparency, and safety. 

• Automate your data preparation and validation abilities to slash preparation time and gain insights into metrics with minimal intervention
• Consolidate data from various sources into a secure environment with guard rails that support compliance, ESG tracking, and operational transparency
• Gain real-time visibility of metrics, to adapting workflows and adjusting operational activities to meet regulatory requirements and sustainability goals
• Enhance the safety of frontline workers with better reporting and documentation of incents providing intelligent insights that optimize your safety culture  

This session explores the use of Generative AI to automate and improve Root Cause Analysis (RCA) on customer quality complaint data. By combining domain-specific language models with structured and unstructured feedback records, Dow’s system uncovers hidden patterns, links operational issues to customer experiences, and generates coherent RCA summaries. The approach streamlines investigation workflows, enhances consistency in issue resolution, and supports proactive quality improvements.

Swee-Teng Chin

Statistics Technical Leader
Dow

See Bio

Learn how AI voice-enabled assistants and smart wearables are empowering field workers with instant access to data, procedures, and remote support, boosting productivity while keeping hands and eyes on the job. 

This session examines methods that combine artificial intelligence with other approaches to improve predictive diagnostics in industrial production, particularly in scenarios with limited data and uncertainty. The discussion includes examples of AI applications for predicting heat exchanger fouling. The case study demonstrates how integrating physics-based models with machine learning can enhance reliability, reduce downtime, and support more cost-effective decision-making in energy management. 

Ivan Castillo

Senior Data Scientist
Dow

See Bio

Powerful simulation tools provide the opportunity to increase the reliability, resilience and productivity of your asset portfolio. By creating models of your assets your organization can model, predict and optimize performance. From understanding the likelihood of equipment failures to drive predictive maintenance, to understanding how your assets will behave in different operating conditions to drive performance optimization, forecasting the health of your asset is key to retaining safe and efficient operations. 

When every function across ExxonMobil came knocking with the same request of “We need a smarter way to interact with our documents”, the data science team knew an Enterprise-scale generative AI powered solution was the answer. The vision was to create a solution that empowered employees to ‘speak with their documents’, instantly summarizing complex technical reports, contracts, and operational manuals into actionable insights. The potential to drive productivity enterprise wide and eliminate hours of manual effort was undeniable.

Jason Gee

Manager, Data Science
ExxonMobil

See Bio

Data alone doesn’t drive optimal decision making, context does. As organizations integrate diverse data sets across business units, advanced analytics and visualization platforms are enabling a new level of operational clarity. In this session, we’ll explore how combining smart energy analytics with human intelligence unlocks deeper insights, improves forecasting accuracy, and supports smarter resource allocation. Discover how energy leaders are using intuitive dashboards and cross-functional data models to make faster, more reliable decisions, while maintaining the human judgment that keeps operations resilient and adaptive.