Marie Merle Caekebeke

Sustainability Executive – Strategic Engagement, Baker Hughes

Supernova Award Category: 
ESG & Sustainability
The Organization: 

Baker Hughes is an energy technology company that provides solutions to energy and industrial customers worldwide. With the complex energy market and challenging macro environment, there is a renewed focus on energy security, sustainability, and affordability. As Baker Hughes restructures and simplifies its business into two segments, Oilfield Services & Equipment, and Industrial & Energy Technology, the vision is to be a sustainable pioneer and weave sustainability more deeply into the corporate culture.

The Problem: 

At Baker Hughes, sustainability means operating in a principled way to minimize environmental impact and maximize social benefits by providing affordable, sustainable, and secure energy for people and the planet. Every two years, Baker Hughes conducts a detailed ESG materiality assessment – an important listening exercise for internal and external stakeholders – to strategically align key sustainability priorities with the highest business impact and stakeholder importance. This biennial assessment helps inform strategic sustainability priorities while providing an opportunity for deep engagement with key stakeholders. While foundational in developing strategic priorities that support the commercial strategy through more sustainable operations, the standard ESG materiality exercise is a labor-intensive process that is inherently limited in scope due to resource constraints. In a world of constantly evolving stakeholder expectations around sustainability, Baker Hughes looked to incorporate a tool to widen the aperture of stakeholder signals and provide continuous and actionable feedback in order to mitigate potential ESG risks and capture opportunities.

The Solution: 

Baker Hughes recognized the power of advances in natural language processing (NLP) and large language models (LLMs) to fill the gap in their existing ESG materiality process. Turning to its AI technology partner C3 AI, they explored ways to execute a more frequent stakeholder ESG materiality assessment across a much broader set of sources than possible with the previous manual methods. Using C3 AI ESG, Baker Hughes was able to, in just 9 weeks:
- Parse over 3,500 stakeholder documents from a broad range of external stakeholders into 400,000+ paragraphs
-Train NLP ML pipelines to identify and label paragraphs aligned with key ESG topics, including 1,700+ training labels
- Deploy a workflow to compute time series ESG materiality scores for source documents at the paragraph, document, stakeholder, and stakeholder group levels
- Configure application interface to visually represent ESG scores, stakeholder analysis, evidence packages, and other benchmarking insights.

The Results: 

By incorporating the C3 AI ESG application into its ESG materiality process, Baker Hughes leveraged AI to review key areas of interest for stakeholders. Baker Hughes used the results of this analysis to advise the development of a sustainability strategy geared towards catalyzing opportunity and buffering risk by focusing on the areas that matter. These insights help the team transition time spent from manually monitoring reams of data sources to driving strategic initiatives that improve ESG performance.

Baker Hughes’s improved ESG materiality approach combines carefully curated NLP techniques to eliminate bias and ensure consistent scoring of incremental stakeholder data. It also uses AI to parse a high volume of stakeholder data in a matter of hours to surface actionable insights enabling the early identification of shifts in stakeholder ESG priorities. Benchmarking stakeholder ESG materiality scores against NLP-perceived Baker Hughes ESG materiality scores enables added visibility to further align on priorities and identify any gaps.

C3 AI’s ESG materiality assessment captures stakeholder emphasis on greenhouse gas emissions, further supporting Baker Hughes’ continued efforts towards its emissions-reduction pledges..


The Baker Hughes deployment of C3 AI ESG dramatically widened the aperture of possible stakeholders, documents, and document types considered for materiality assessment. Baker Hughes ingested and reviewed data from 3500 documents from a broad range of stakeholders such as investors, customers, competitors, NGOs, and the company itself.

Across these stakeholders, ESG expectations are communicated through various documents, such as investment engagement guides, press releases, and reports, which can easily add up to 26,000 pages a year. A human takes 790 hours to analyze that volume of content, while the C3 AI ESG application takes less than one hour to analyze and incorporate those insights. Further, only 10% of the text contained relevant content, tedious work for a human to process but easy for a properly configured NLP pipeline.

There are significant labor efficiency benefits of automating the AI-driven assessment of over 400,000 paragraphs of text. Applying a manual process to this amount of data would require almost 30,000 hours and a 2-year cycle time to complete the materiality assessment. Baker Hughes compared the AI NLP results against the outputs of their recent manual materiality assessment and found widespread alignment in ESG topic materiality scores, substantiating confidence in the ongoing AI-produced outputs.

The Technology: 

C3 AI ESG enables companies to manage and improve their ESG performance with advanced machine learning, natural language processing and generative AI techniques—significantly reducing time and resource constraints. In addition to the materiality assessment module, customers can utilize C3 AI ESG to unify and store disparate ESG data as ESGbitsTM, automate reporting to standards, boost specificity and traceability of carbon emissions calculations, and manage ESG plans with scenario analysis.

Disruptive Factor: 

The C3 AI ESG product provides, to our awareness, the first targeted ESG stakeholder materiality assessment fully automated using AI NLP. With its proprietary NLP ensemble model methodology, continuous actionable AI-insights, and rapid deployment, this application is unique.

Baker Hughes’ manual ESG materiality assessment process is a manual exercise executed once every two years. Leveraging AI enables more frequent ESG topic reviews while significantly reducing time and resource constraints. in a landscape of ever-evolving stakeholder expectations. With C3 AI ESG, Baker Hughes can incorporate materiality signals from a much broader set of stakeholders and inputs. Ultimately, the company can adapt more quickly and proactively to shifts in specific priorities; the application goes beyond an automated materiality assessment to provide a stakeholder intelligence tool.

Shining Moment: 

In addition to external stakeholders, the NLP pipeline processed internal Baker Hughes data to generate intrinsic insights on how the company is perceived to be communicating and prioritizing ESG issues. Benchmarking the perceived Baker Hughes scores with its competitors’ uncovered valuable opportunities to leverage stakeholder feedback to drive strategic goals and performance.

About Baker Hughes

Baker Hughes (NASDAQ: BKR) is an energy technology company that provides solutions to energy and industrial customers worldwide. Built on a century of experience and conducting business in over 120 countries, our innovative technologies and services are taking energy forward – making it safer, cleaner and more efficient for people and the planet. Visit us at