For large pharmaceutical companies, regulatory work often depends as much on what has already been done as what needs to be done next. Every interaction with a health authority adds to a growing body of knowledge. Over time, that knowledge becomes an asset, if the company can access and apply it.
That was a challenge at one global pharma company. Years of submissions and correspondence had created a vast repository of documents, emails, formal responses, and supporting materials stored within its regulatory systems. The content was valuable but difficult to use in practice.
When new questions came in from health authorities, teams often had to start fresh. Relevant precedent might exist somewhere in the system, but finding it wasn’t easy. Search tools were available, but they were difficult to navigate, especially for less experienced staff.
The company had a seemingly simple question: How can we make this information easier to find and use?
IPM helped the company in answering the question through a focused, cross-functional effort. Regulatory experts, technical teams, and stakeholders across regions came together to define what information mattered most and how they would use it day to day.
One of the first hurdles was access. The information existed across millions of documents within the regulatory information management (RIM) system. Extracting the data and making it usable outside of the source system required both technical effort and vendor management.
Similarly, the team needed to resolve questions about data sensitivity and access before the work could move forward. While the RIM system controlled access at the document, person, or country level, this security would be lost once the documents were extracted from the proprietary system. Not everyone who would use the data outside of RIM currently had clearance to see it. Decision-making required communication up and down the executive ladder, laying out benefits and risks and addressing concerns. Ultimately, senior leadership decided that the benefit of open access outweighed the risks for most documents.
Once the data was available, the team focused on how to work with it in a practical way. They used AI tools to parse large volumes of content to isolate health authority questions and the corresponding responses, then organize those elements into a structured database. This made it possible to search for similar questions and see how the company had responded in the past. Regulatory scientists could reuse prior work more effectively and reduce the time spent recreating responses while ensuring consistent responses to agencies.
As the work progressed, the team extended the concept. Instead of limiting the solution to specific document elements, they explored how to make the broader body of regulatory information accessible. They built an AI model to ingest the full set of documents and provide information via a chatbot. Users could ask questions more naturally, without relying on predefined tags, categories, or specific search criteria.
The evolution was an expansion of the same objective: improve access to information in a way that supports how people actually work.
Regulatory teams can now search for and retrieve relevant information more efficiently, which reduces effort and shortens response time for health authority inquiries. Just as important, they have greater visibility into prior responses, which helps maintain consistency across submissions and over time.
The work also surfaced broader considerations that extend beyond a single application. By focusing on a defined objective—making regulatory knowledge easier to access and apply—the organization created a foundation that can continue to evolve as technology advances. AI played a meaningful role in that progress, not as the starting point, but as a way to make the underlying objective achievable at scale.
Future enhancements include feeding past relevant responses into an AI model that will create a first draft ready for human review.