Oil & Gas Leader Boosts Employee Efficiency with GenAI Search

A major downstream oil and gas corporation leveraged generative AI to improve operational efficiency by making internal documents, architectural standards and best practices easily searchable in a conversational way. This initiative not only saved users significant time but also boosted overall productivity and standardization across the organization.

This downstream oil and gas company managed its IT architecture documents within a 200GB repository on Azure-hosted Microsoft SharePoint servers. Although the infrastructure was robust, locating specific information within this repository could often be tedious and challenging for users who may not know exactly where to look for it. 

To address these issues, Publicis Sapient partnered with the client to revolutionize data accessibility by incorporating advanced generative AI and AI solutions. This technology facilitates precise data extraction and provides a seamless presentation of information through an innovative conversational AI interface designed for intuitive user interaction. This strategic enhancement aims not only to improve efficiency but also to transform the user experience by simplifying access to critical data and reducing the complexity of navigating the repository.

The initiative ensures the secure storage of SharePoint data on Azure. Using AWS Amplify, AWS Fargate, AWS Lambda and either Amazon Kendra or Azure Cognitive Search, users can enter their queries in a chatbox on the frontend and receive a LLM-generated response based on documents pulled from a repository.

By seamlessly integrating with AWS Gen AI, this solution transforms information retrieval, offering unparalleled efficiency and accuracy. This innovative approach not only streamlines information access but also establishes a direct link to the source, ensuring a comprehensive and seamless user experience through an intuitive web-based interface.

By modernizing enterprise search with generative AI, the organization fundamentally changed how teams access and use critical knowledge. What was once a time-consuming process of navigating massive document repositories became a fast, conversational experience that delivers accurate, summarized answers in seconds.

Beyond immediate efficiency gains, the initiative accelerated the organization’s broader AI ambitions—informing infrastructure decisions, validating large language models and helping establish a Generative AI Center of Excellence to scale innovation responsibly across the business

The results were substantial and measurable:

• Near-instant access to knowledge: Average query time dropped to around 20 seconds, down from roughly five minutes.
• Enterprise-scale natural language search: More than 200GB of documents are now fully searchable through conversational AI.
• Dramatic productivity gains: Productivity increased by 93 percent by eliminating manual search and extraction.
• Higher accuracy and consistency: Data retrieval accuracy improved by 94 percent, while standardization across programs increased by 96 percent.
• Stronger AI foundations: Evaluation of multiple LLMs and indexers informed optimal architecture choices and advanced enterprise-wide AI maturity.
  

Together, these outcomes show how generative AI can move beyond experimentation to become a practical, scalable driver of operational efficiency and standardization.