Australian resources company Woodside will use IBM's 'cognitive computing' system Watson to extract historical data from project reports.
The company is hoping Watson will help its engineers better design, fabricate and construct oil and gas facilities.
"We are bringing a new toolkit to the company in the form of evidence based predictive data science that will bring down costs and increase efficiencies across our organisation," Woodside senior vice-president, strategy, science and technology, Shaun Gregory said in a statement.
"Data science, underpinned by an exponentially increasing volume and variety of data and the rapidly decreasing cost of computing, is likely to be a major disruptive technology in our industry over the next decade.
"Our plan is to turn all of this data into a predictive tool where every part of our organisation will be able to make decisions based on informed and accurate insights."
The cloud-based Lesson Learned service will allow Woodside engineers to make natural language queries to extract data from three decades' worth of project reports.
Last month the Department of Immigration and Border Protection revealed it would use IBM Watson platform to analyse unstructured data.
The department said it would use the Watson to analyse data drawn from sources including news feeds and government reports.
"We are hoping that Watson will allow us to more effectively manage the information overload problem by detecting signals in the very noisy world of unstructured, open source data," a department spokesperson said.