Using Generative AI to Revolutionize Compliance Queries in the Energy Sector: A Case Study from Cepsa Química and Keepler
Generative artificial intelligence (AI) is revolutionizing businesses across various industries, offering a way to streamline processes, generate human-like content, and drive innovation. The energy sector is not exempt from this paradigm shift, as generative AI can bring substantial value in terms of efficiency and productivity.
Cepsa Química, a leading company in the manufacturing of chemical products, recognized the potential of generative AI in enhancing their product stewardship processes and partnered with Keepler, a cloud-centered data services consulting company, to implement a generative AI assistant. This assistant aims to expedite compliance queries related to the chemical products they market, helping the team save time and improve overall operational efficiency.
The partnership leveraged Amazon Bedrock, a fully managed service that offers high-performing foundation models from leading AI companies, to build the generative AI solution. By using a Retrieval Augmented Generation (RAG) approach, the team ensured that the AI assistant could dynamically adapt to changes in regulatory information, providing up-to-date responses to user queries without the need for retraining.
The solution developed by Cepsa Química and Keepler is based on four main functional blocks: input processing, embeddings generation, LLM chain service, and user interface. It is divided into two modules, one for batch processing input documents and another for answering user queries through inference. These modules work together seamlessly to provide quick and accurate responses to compliance queries.
Throughout the development process, the team faced several challenges, such as data preprocessing complexities and evaluating the results of the AI models. To address these challenges, they implemented strategies like data chunking, model selection, and query variants, resulting in significant improvements in retrieval and response accuracy.
The implementation of the generative AI assistant has led to various improvements for the product stewardship team, including faster query times, enhanced answer quality, and increased operational efficiency. Moving forward, Cepsa Química plans to identify additional use cases for generative AI across different business functions, aiming to create a corporate-wide tool that leverages the success of their initial initiative.
In conclusion, the collaboration between Cepsa Química and Keepler showcases the potential of generative AI in transforming operational processes and driving efficiency in the energy sector. By harnessing the power of AI technologies like Amazon Bedrock and RAG techniques, businesses can unlock new opportunities for innovation and productivity. If you’re interested in integrating generative AI into your business, reach out to specialists in the field or explore platforms like PartyRock to kickstart your AI journey.