Waters’ new AI tool targets one of the pharmaceutical industry’s most tedious tasks

Brook Colangelo, chief information officer at Waters Corp., says the lab equipment company’s newest artificial intelligence tool can help drugmakers speed up their quality testing processes.

On Wednesday, Waters debuted a new AI-powered integration called Peak Integration into its Empower software platform to drastically reduce manual interventions traditionally performed by scientists for quality assurance (QA) and quality control (QC) in drug discovery and manufacturing. These QA/QC processes ensure medicines are safe, effective, and compliant with the Food and Drug Administration and other regulatory standards.

Waters says that the new AI and machine learning capabilities within Empower—which has 500,000 users and supports the release of popular drugs ranging from aspirin to Wegovy—can reduce manual interventions by 95%. Waters also says that the new AI feature improves testing accuracy by 76% and is 50% faster than the prior process.

“The advancements here are around how we take automation and how we take AI to help unlock more human potential in the scientists that are using this technology,” says Colangelo, who has served as CIO at Waters since 2018.

The latest AI/ML innovation from Waters required a bit of patience. Waters worked closely with an undisclosed, top-ten pharmaceutical company for about one-and-a-half years, with much of that time dedicated to securing legal agreements and guaranteeing that there were safe processes in place to exchange data to train the initial AI models.

“Once we got into it and had a novel model that we started training, it actually went pretty quickly from there,” says David Leitham, vice president and general manager of Waters’ informatics software business.

It is the second major AI feature that Waters has announced for Empower, following an AI/ML anomaly detection tool that debuted in February 2025, and also the second major AI announcement from the company in October. Two weeks ago, Waters launched “AI Link,” a software capability that brings Anthropic’s Model Context Protocol—the AI startup’s open standard that connects AI models and agents to software and data through a common interface—to the lab company’s informatics platform.

“We have been less vocal than some of our competitors,” says Colangelo, explaining Waters’ current efforts to laud its AI capabilities.

Leitham says that Waters’ new AI feature, called Peak Integration, targets repetitive, mundane tasks that are “frankly not a good use of the scientist’s time or expertise, but also add variability to the results of the experiment.” A scientist could be tasked with reviewing hundreds of chromatograms, graphs that show separate compounds in a mixture or formulation, but would need to hunt for just one or two that don’t fit the desired profile.

While the general public’s view of AI is shaded by concerns about its impact on jobs, the environment, and even humanity’s future—so much so that President Donald Trump wants to ban the term artificial intelligence in favor of his preferred name, super intelligence—drug discovery is one AI use case that occupies more favorable terrain.

There are hopes that AI can accelerate drug discovery and reduce the $2 trillion, 10-year process of getting medicines approved and into the market, improving everything from research to development to clinical trials and manufacturing, as well as targeting medicines to the most appropriate patients.

Pharmaceutical giants have aggressively announced partnerships with major AI players, including deals between Nvidia and Eli Lilly, Novo Nordisk and OpenAI, and Merck and Alphabet’s Google Cloud, all among the more than three dozen pacts that have been announced throughout 2026, according to Reuters.

“There have been a lot of pharma and life sciences companies that have talked a lot about AI, but not in this area,” says Colangelo of the new Waters’ tool and its focus on QA/QC.

Internally, Colangelo’s AI bets have included giving more than 2,000 Waters employees access to the conversational tool called AI Advisor. In its manufacturing operations, Waters built a scientific data lake to unify historical and current engineering, quality, and production data. AI added to the manufacturing processes provides greater end-to-end traceability and has higher-quality production, Waters says.

Another way Colangelo intends to rely on more AI usage involves the integration of a biosciences and diagnostics solutions business that Waters bought from medical technology firm Becton Dickinson. That deal—which more than doubled the size of Waters—valued the business at $18.8 billion when the deal closed in February.

He says that the Becton Dickinson unit had 8,500 IT employees, though only around 20 joined Waters. And very few employees from other core functions, like human resources and finance, also migrated to Waters following the transaction.

“We have to replace a lot of that work,” says Colangelo, who says he intends to generate efficiencies from AI, though he couldn’t share concrete details at this time. “Instead of doubling the size of the IT team, we’re going to use AI to transform it.”

John Kell

DON’T MISS: Cyber threats don’t stand still—and neither do the startups working to stop them. From AI-powered deception to hidden vulnerabilities, businesses face an increasingly complex security landscape. For the fourth annual Cyber 60, Fortune, Lightspeed Venture Partners, and AWS spotlight the most innovative cybersecurity startups helping companies anticipate threats, strengthen defenses, and protect their operations.

This story was originally featured on Fortune.com