Business 5 min read By Alice Ashford
Waters launches AI tool to cut manual steps in drug quality testing
Waters Corp. has added an AI-powered feature to its Empower software that it says reduces manual intervention in pharmaceutical quality assurance and quality control by 95 per cent.
Waters Corp., the lab equipment company, has launched an artificial intelligence tool designed to speed up quality testing for drugmakers. The feature, called Peak Integration, was introduced on Wednesday as part of the company's Empower software platform and aims to cut down on the manual work scientists have traditionally carried out in quality assurance and quality control during drug discovery and manufacturing.
Quality assurance and quality control processes are central to ensuring that medicines are safe, effective and compliant with the US Food and Drug Administration and other regulatory standards. Waters says the new AI and machine learning capabilities within Empower, which has 500,000 users and supports the release of drugs ranging from aspirin to Wegovy, can reduce manual interventions by 95 per cent. The company also says the feature improves testing accuracy by 76 per cent and is 50 per cent faster than the previous process.
Brook Colangelo, chief information officer at Waters, who has held the role since 2018, said the goal is to use automation and AI to unlock more human potential among the scientists using the technology. David Leitham, vice president and general manager of Waters' informatics software business, described the repetitive tasks the tool targets as frankly not a good use of a scientist's time or expertise, adding that they also introduce variability into experimental results.
He gave the example of a scientist who might be asked to review hundreds of chromatograms, graphs showing separate compounds within a mixture or formulation, in order to find the one or two that do not fit the desired profile. Peak Integration is designed to take over much of that screening.
The development took some time. Waters worked for about a year and a half with an undisclosed top-ten pharmaceutical company, much of which was spent securing legal agreements and putting in place safe processes for exchanging the data used to train the initial AI models. Once the team had a novel model in training, Leitham said, progress moved quickly.
Peak Integration is the second major AI feature Waters has announced for Empower, following an AI and machine learning anomaly detection tool introduced in February 2025. It is also the second significant AI announcement from the company this month. Two weeks earlier, Waters launched AI Link, which brings Anthropic's Model Context Protocol, an open standard that connects AI models and agents to software and data through a common interface, to its informatics platform. Colangelo acknowledged that the company has been less vocal about its AI work than some competitors, but said it is now stepping up its communication.
The move comes as pharmaceutical companies increasingly look to AI to shorten drug development, a process that involves roughly $2 trillion and about ten years before a medicine reaches the market. Large drugmakers have announced a series of partnerships with major AI players this year, including deals between Nvidia and Eli Lilly, Novo Nordisk and OpenAI, and Merck and Alphabet's Google Cloud. Colangelo said many pharmaceutical and life sciences firms have talked extensively about AI but few have focused on quality assurance and quality control in the way Waters has.
Waters has also expanded its own use of AI internally. More than 2,000 employees have access to a conversational tool called AI Advisor, and in manufacturing the company has built a scientific data lake that unifies historical and current engineering, quality and production data. Waters says AI applied to manufacturing gives greater end-to-end traceability and higher-quality production.
The company's AI plans are also tied to its acquisition of a biosciences and diagnostics business from medical technology firm Becton Dickinson. The deal, which more than doubled the size of Waters, valued the business at $18.8 billion when it closed in February. Colangelo said the unit had 8,500 IT employees but only around 20 joined Waters, and very few staff from other core functions such as human resources and finance also transferred. He said the company must replace much of that work and intends to generate efficiencies through AI, though he did not share concrete details. Instead of doubling the size of the IT team, he said, Waters plans to use AI to transform it.
3
