Microsoft & Novo Nordisk: Accelerating Pharma R&D With AI

Rising R&D costs and intensifying competition require pharmaceutical firms to extract value rapidly from legacy data. Historically, answering scientific questions in R&D demanded complex coordination across medical experts, commercial stakeholders, data scientists and biostatisticians. Teams spent months evaluating hypotheses, frequently finding insufficient evidence to proceed.

Through its FounData initiative, Novo Nordisk has harmonised more than 200,000 patient-years of clinical trial data into a cloud platform. The data aligns with open standards, enabling immediate analysis without bespoke pre-processing.

However, data access still depends on scarce specialist expertise, delaying workflows and trial timelines. To resolve this, Novo Nordisk sought to accelerate R&D decision-making by helping researchers explore and validate clinical hypotheses faster using AI-assisted quantitative analysis.

Reasoning agent architecture

Collaborating with Microsoft's Forward Deployed Engineering team, Novo Nordisk developed a governed quantitative decision-support agent on Azure. The solution enables experts to independently explore hypotheses against proprietary datasets using AI-generated code execution and statistical analysis.

Unlike chatbots that simply retrieve answers, this agent generates code, navigates structured clinical datasets and explores statistical workflows to deliver explainable outputs. Researchers can input complex scientific questions in natural language and receive data-driven analyses in minutes rather than weeks.

“We wanted to move from gut-feel decision-making toward quantitative decision support. If we can validate ideas earlier, fail faster when necessary, and prioritise stronger opportunities sooner, that changes the economics of pharmaceutical development,” says Sid Prabhu, Senior Director, Head of FounData AI Application at Novo Nordisk.

The system's reasoning capabilities exceeded expectations.

"What surprised people internally was not just that the system could generate answers," Sid notes. "It could reason through problems, identify confounding factors, and surface insights in ways that made sense to scientific experts."

Governance and operational trust

The collaboration focused on enterprise deployment within a regulated research environment. Microsoft provided the infrastructure, governance patterns and scalable orchestration to deploy AI safely against proprietary clinical data.

"You don't just connect an LLM to enterprise data and hope it works," Sid explains. "There's a tremendous amount of infrastructure, governance, testing, and evaluation required to make something like this production-ready in a regulated environment."

Prior to deployment, the teams executed thousands of automated tests and implemented layered evaluation systems. Human-in-the-loop workflows allow biostatisticians and subject matter experts to validate outputs before they influence major scientific decisions, resolving early internal skepticism.

"The agent is there to inform decisions, not to take over control. It drafts the analysis; the scientist decides what to do with it. That distinction is foundational to how we built it," says Rasmus S. Andersen, Associate Director, Products and AI, FounData at Novo Nordisk.

Business impact and future roadmap

The solution reduces time to insight from weeks to minutes, increasing analytical throughput and evaluating more high-value opportunities earlier.

"In the past, we might have had the capacity to pursuefive to 10 strong ideas in a quarter," says Sid. "Now we can evaluate more than 50."

The system augments human expertise rather than replacing it, eliminating low-value exploratory tasks so specialists can focus on critical priorities. "The goal was never to replace scientific expertise," Sid says. "The goal was to help our experts spend more time on the decisions that matter most."

Mishal Patel, GVP AI & Digital Innovation at Novo Nordisk, outlines the broader benefit: “What's powerful here isn't speed alone – it's that we can now explore far more scientific questions earlier, with greater productivity, efficiency, and scientific quality while maintaining the rigor our work demands. Ultimately, that increases our probability of success at the very front end of drug development.”

Novo Nordisk is currently expanding the tool into trial design optimisation and portfolio intelligence. The future roadmap includes shifting toward predictive capabilities, such as outcome modelling and trial-scenario simulation, while expanding to broader data modalities like omics, device telemetry and external scientific literature.