How AI Accelerates Oil Production and Saves Thousands of Hours of Support
In the News
How AI Accelerates Oil Production and Saves Thousands of Hours of Support
The American energy technology company Baker Hughes, together with EPAM and Amazon Web Services (AWS), developed two digital assistants based on generative AI (GenAI). One supports production engineers, the other helps service teams in the field. The tools provide immediate time savings. The creators also see them as a starting point for wider use of GenAI within the energy sector.
A team from EPAM, a digital transformation and product engineering company, built a working prototype in four weeks. The assistant immediately retrieves production data, recognizes deviations and advises on next steps. It generates SQL queries, creates interactive visualizations, and runs entirely on AWS, including, Amazon Bedrock and Anthropic's Claude 3.5 model.
85% technical questions answered correctly
Baker Hughes' Artificial Lift Systems (ALS) division answers around 1,600 technical questions from field technicians worldwide every year. Many of these questions come up more often. Experts spend an estimated 10,000 hours on this every year. To reduce workload, Baker Hughes and EPAM developed a GenAI assistant that handles these repetitive questions independently. The team used Amazon Textract, Titan models, and OpenSearch to train the assistant with knowledge from existing documents such as manuals, technical updates, and procedures. After extensive testing, the assistant now answers 85 percent of the questions correctly. The next step is to expand with more documents. Integration into the ALS portal is also planned. The goal is for technicians worldwide to use the assistant in the long term.
Read the full article here.
Learn more about EPAM’s partnership with Baker Hughes: https://www.epam.com/services/client-work/leveraging-genai-to-increase-productivity
