Skip to content
What We Do
Climate Corps Projects

SKF – Xinyuan Xu – 2026

The project addressed fragmented ESG information, repeated manual searches, limited reuse of approved responses, and difficulties in identifying the right internal contacts for specific customer questions. It consolidated public and internal ESG materials and historical questionnaires into a structured knowledge base and built an ESG AI Agent to support information retrieval and questionnaire responses, improving information accessibility, response consistency, and overall completion efficiency.

The project focused on challenges in ESG questionnaire responses, including fragmented information sources, repeated manual searches, and limited reuse of previously approved responses. As customer ESG questionnaires increase in volume and complexity, employees often spend considerable time locating information and coordinating across departments. The project therefore aimed to consolidate public company materials, historical ESG questionnaires, and internal resources into a structured knowledge base and build an ESG AI Agent to provide traceable draft responses and suggest relevant departments or contacts for questions requiring further input, improving information accessibility, response consistency, and overall questionnaire efficiency.

The project first collected, organized, and categorized a range of ESG-related materials, including historical customer ESG questionnaires, annual reports, sustainability reports, company policy documents, and internal reference materials. These resources were then structured based on key attributes such as ESG topic, question type, source document, and approved responses to create a reusable ESG knowledge base. Building on this foundation, an ESG AI Agent was built to support questionnaire-related information retrieval and response generation. By combining large language models (LLMs) with Retrieval-Augmented Generation (RAG), the Agent can retrieve relevant content from the knowledge base, generate draft responses, and provide source references for subsequent review.

The project improved ESG knowledge reuse and streamlined the questionnaire response process by turning dispersed information into a structured and accessible knowledge resource. The ESG AI Agent enables users to locate relevant information more efficiently, generate draft responses, and trace answers back to their sources, reducing repeated manual searches and supporting greater consistency across questionnaires. Together, the knowledge base and AI Agent provide a foundation for more scalable ESG information management and questionnaire response. Ongoing updates and review by relevant departments will remain important for maintaining the accuracy, reliability, and applicability of generated responses as ESG requirements evolve.

At a glance
Industry: Industrial Goods and Manufacturing
Project types:
  • Data Analysis
  • Sustainability and Energy Management Strategy
Year: 2026
Location: Shanghai, China
About the fellow
Xinyuan Xu

Who We Work With

EDF has collaborated with over 40% of Fortune 100 companies to align sustainability goals with bottom line gains