Kuken Industries, a maker of cooling towers, now uses AI to dig through its old drawings. Engineers can find the drawing they need for a repair in 4.5 minutes. It used to take nearly an hour. The company uses the OpenAI API to search an archive that covers 70 years of cooling tower designs.
What changed
Cooling towers are large structures that get rid of heat from buildings and industrial sites. They last a long time, so repairs often call for drawings made decades ago. Finding the right one in a huge archive is slow work. Kuken now runs that search through the OpenAI API, which is a tool that lets a company build OpenAI's AI into its own software.
Who it affects
The direct users are Kuken's engineers, who spend less time hunting for files. The story also points to a wider group. Any business that keeps decades of paper or scanned records, such as plans, contracts or service logs, faces the same kind of slow search.
What is not known yet
The material gives few details. It does not say how the system was built, what it cost, or how many engineers use it. It does not explain how the 4.5 minute figure was measured, or whether results are checked by a person before a repair. The full article was not available, so those points remain open.
What to do
A business with a large archive can start by timing a typical search today. That gives a baseline to compare against any new tool. Kuken's numbers come from the company's own account, so treat them as one example, not a promise for every archive.
The takeawayIf your staff lose hours finding old records, time a typical search now so you can judge whether AI search is worth testing.