Speeding up pharmaceutical research through automated data extraction
Branding Science
We accelerated pharmaceutical research projects by unlocking market insights from two decades of data.
Increase in data sources from the initial project phase
Projects can now be queried quickly and easily
Background
Over the last 20 years, Branding Science has carried out more than 1,200 projects in the past decade for its pharmaceutical clients. However, their qualitative research was confined to its original output files, leading to siloed and underutilised insights. If they could quickly surface two decades’ worth of market insights on rare disease and oncology, it could speed up the development of vital drugs.
“By using the latest machine learning techniques to uncover qualitative insights, this project marks a step change in market research.”
Ben Walker, Engagement Manager
Faculty
Solution
We developed a platform that integrates large language models (LLMs) with Branding Science’s internal knowledge base. The solution uses search and retrieval functionality to enable Branding Science staff to quickly look-up relevant projects and interviews. It uses semantic search so users can query their extensive knowledge base using natural language through an accessible interface.
Impact
Branding Science’s teams can now access qualitative insights from over 20 years worth of data. This approach has the potential to transform pharmaceutical research, market analysis, and data practices.
“This is not just a testament to our ongoing commitment to innovation, but also to making a meaningful difference to patients' lives by empowering healthcare providers with superior tools.”