How ENEOS Materials cut HR analysis time by 90% with ChatGPT

The manufacturing sector is often painted as a late adopter, but ENEOS Materials just made a strong case that the opposite can be true. The Japanese company, a core entity of the ENEOS Group that makes everything from tire rubber to battery binders, rolled out ChatGPT Enterprise across its entire workforce and is now reporting the kind of numbers that make IT departments pay attention.

During the pilot, 80% of employees reported significant workflow improvements. That’s a strong signal, but the HR department’s result is the real standout: a 90% reduction in data aggregation and analysis time. When a cost-center function sees that level of efficiency gain, it makes the ROI conversation for the C-suite a lot simpler. The company has since created over 1,000 custom GPTs, with more than 90% of employees using the tool at least weekly.

Some of the most dramatic time savings came from the unexpected application of deep research. Kenichi Sakemi from the Process Development and Engineering Department pointed to a plant in Hungary where scouring local-language sources used to take months. “What once took months… now takes tens of minutes,” he said, noting that ChatGPT Enterprise can comprehensively search local materials and spit back precise Japanese translations. For highly specialized domains like chemical engineering, calculations that consumed half a day are now finished in minutes, simply by asking questions in Japanese.

The engineering team is also using custom GPTs to design plants based on internal standards, generating optimized specs from inputs like fluid type and pipe diameter in seconds rather than hours. Beyond speed, the tool flags material-selection risks, simultaneously tightening safety standards and cost efficiency. Yoshirou Sakura, a manager in the Production Technology Department, framed the adoption bluntly: using digital tools to boost productivity is essential as the workforce shrinks. For a manufacturing industry grappling with a declining birthrate and rising costs, ENEOS Materials’ results show that secure, proprietary-data-friendly AI isn’t just a productivity hack — it’s becoming a competitive requirement.