Enhancing ESG Reporting with Knowledge Graphs and LLMs for Precision and Scale
A Talk by Adam Wangrat (Lead Knowledge Graph Engineer, Neural Alpha)
About this Talk
Financial professionals face an overwhelming amount of unstructured data when preparing reports across multiple ESG-related frameworks, such as ISSB, TCFD, and TNFD. The challenge lies in efficiently identifying material risks and opportunities from vast data sources, especially when these involve critical environmental and sustainability concerns.
This talk will share universal lessons about combining knowledge graphs with AI, offering attendees practical insights they can apply to their challenges. Moreover, it highlights nature and biodiversity considerations, which are increasingly relevant in ESG reporting and investment decision-making.
The session will appeal to those already working with knowledge graphs and large language models (LLMs), as well as those interested in integrating these technologies. Specifically, it targets professionals working with GraphRAG architectures or looking to ground LLM workflows in knowledge graph data.
To tackle the challenge of unstructured data in ESG-related reporting, we built a solution that combines the power of knowledge graphs and large language models (LLMs). We structured ESG indicators in a knowledge graph, allowing us to map material risks to specific business activities across multiple frameworks. LLMs were employed to identify which indicators were relevant based on the business sector, such as water usage being critical for certain industries but not others.
Our system uses this contextualized data to extract material indicators from company disclosure documents, ensuring only the relevant information is included. The extraction process is further supported by our ESG taxonomy, which links to the indicators, providing additional context and improving the precision of the data retrieval. This method enables us to automatically generate portfolio-level reports, significantly streamlining the process for financial professionals while ensuring accuracy in identifying risks and opportunities.
Although the project is still in development, the expected impact includes significant time savings for financial institutions and improved risk management. This solution has the potential to reduce operational risks while identifying opportunities for nature-positive investments.
About The Speakers
Lead Knowledge Graph Engineer, Neural Alpha
Adam drives the development of innovative solutions integrating ESG data into knowledge graphs. With expertise in AI and graph technologies, Adam focuses on creating scalable, accurate systems for responsible investment. He holds an MSci in Computer Science (AI) from King’s College London.
We’re a cloud tech company that provides organisations around the world with computing infrastructure and software to help them innovate, unlock efficiencies and become more effective. We also created the world’s first – and only – autonomous database to help organise and secure our customers’ data.
