Case Study:
Automated change requests in dossiers
using LLMs and NLP
How AI is helping regulatory teams increase productivity and reduce guidance monitoring time
Even when you have completed your dossier, new guidances come in with changes that must be reconciled with your dossier. This was a time consuming and people-intensive process – until now.
The Client
The client was spending significant time keeping up to date with monitoring and evaluating new guidances from WHO and others depending on the countries. This was a very resource-intense activity and they wanted a way to automate part of this so that the human in the loop could check it but not do the time-consuming work of actually looking everything up and reconciling the documents.
The Solution
- Combining the use of a specifically-trained LLM (large language model) and another semantic NLP approach to understand text in guidance and identify within guidance documents what text must be changed in the dossier.
- Classifying change level (minor, medium, major and implementing and highlighting the changes in both documents for approval.
The Outcome
- Saves regulatory team vast amounts of time as previously each new guidance had to be read and compared with the dossier but now the changes are classified and suggestions created and the regulatory person simply needs to check these and approve.
To achieve these kinds of results, contact Eularis today.
Latest News
Read our latest blogs here.

Competing and Winning Using AI in Healthcare
I keep finding disruptive AI that solves healthcare marketers’ problems and delivers real ROI in the solutions. But the companies offering them have only 1

AI-Powered Transformation in Pharma Regulatory Affairs
One of the biggest hidden tax on pharmaceutical innovation today is the burden of global regulatory drag. We work in a world where a single

Why Companies That Wait for AI to Mature Are Falling Behind
AI is not just transforming industries; it is also transforming the way we compete. What used to be a luxury is now table stakes, forcing