Waters Corporation launched AI-Powered Peak Integration on October 28, 2026, a new premium capability for its Empower Software that uses artificial intelligence and machine learning to automate one of the most manual and variable steps in chromatography: integrating the peaks in a chromatogram. In a pilot with a large pharmaceutical customer, side-by-side review of 400 chromatograms cut manual interventions from 90 to 4, a 95 percent reduction, while user testing showed 50 percent faster review and 72 percent lower variability between analysts.
The Most Manual Step In The Lab
Manual peak integration is the largest source of variability in chromatographic results and one of the most time-consuming steps in data review. Analysts must decide where peaks begin and end, remove noise peaks, combine high molecular weight peaks and edit baselines, and different analysts frequently reach different answers on the same data. That variability matters because chromatography underpins quality testing for medicines: the results determine whether a batch is released, and inconsistent integration undermines confidence in the data.
Waters’ AI and machine learning-powered integration automates this step inside Empower Software, using models trained on each laboratory’s own data. The system removes noise peaks, combines high molecular weight peaks and edits baselines, producing explainable outputs that analysts can review quickly and with confidence. Version-controlled, no-drift models maintain consistent performance until the laboratory chooses to retrain and approve a new model, preserving the control that regulated environments demand.
Built Inside The Validated Environment
The strategic point is where the AI runs. Empower Software is the chromatography data system of record for regulated laboratories worldwide, used by 500,000 users who support the release of over 80 percent of medicines globally, according to Waters. Because the new capability is built directly into Empower, it operates inside the validated environment where highly protected customer data already resides, bringing AI to the data rather than moving data to the AI, with no compromise to integrity, traceability or compliance, the company says.
“Waters is using AI to help customers bring drugs to market faster and to increase the quality testing of manufactured molecules,” said Tina Wu, senior vice president of Waters Analytical Sciences. She said customers want the speed, accuracy and productivity that AI brings without giving up control of their data or the compliant audit trail behind it, which is why bringing AI inside the validated workflow is a central pillar of the company’s strategy for Empower as a connected, intelligent platform for regulated science.
From Anomaly Detection To Peak Integration
Peak integration is the latest step in a deliberate sequence. The first was AI and machine learning Anomaly Detection, launched in Empower Software in February 2025, which compares target results across chromatography data, surfaces deviations for review, and explains the attributes contributing to each anomaly. Waters is also collaborating with large pharmaceutical customers and academic institutions on autonomous, AI-enabled research, development and quality capabilities that follow the molecule from discovery to release, and is extending the same approach into biologics and bioanalytical characterisation.
AI-Powered Peak Integration is a new premium tier capability available on subscription. Waters is enrolling customers now for early access beginning in December 2026, with validated, compliant deployment to follow in mid-2027. The pilot figures, it should be noted, come from Waters’ own internal pilot study and customer evaluation, not from an independent trial, so laboratories will want to see the same reductions on their own data before treating them as typical.
Why It Matters
This matters because regulated science is where AI faces its hardest adoption test: not whether a model works, but whether its output can be audited, reproduced and defended to a regulator. Waters’ answer is to keep the AI inside the validated system, with explainable outputs and version-controlled models, rather than bolting a clever tool onto the side of a compliant workflow. The same discipline is reshaping virtual patients for drug trials and unified AI platforms for drug discovery, where the promise is faster development but the gate is always evidence a regulator will accept.
The caution is equally clear. A 95 percent reduction in manual interventions in one pilot is a striking result, but it is a company-reported figure from a single large pharmaceutical partner, and chromatography data varies widely between laboratories, molecules and methods. If the models hold up across that variety, reviewers gain time for the judgements that actually need human expertise, much as AI assistance in drug discovery only earns its place by surviving expert review. The mid-2027 validated deployment will be the real test.
Background
Waters Corporation, based in Milford, Massachusetts, and listed on the New York Stock Exchange, is a global leader in life sciences and diagnostics, with about 15,000 colleagues, and its Empower Software is the chromatography data system of record for regulated laboratories worldwide.
Source: Waters announcement
0 Comments