Research automation and clinical literature tooling at Pfizer
A self-updating research database and a pipeline summarising thousands of papers into per-country clinical briefings.
- 2024
- Research coordinator
- Completed
The situation
Medical Affairs across 160-plus emerging markets generates more published literature than any team can read. The work of finding which countries share an epidemiological profile — and therefore which clinical evidence transfers between them — was manual, and repeated every time a question was asked.
What the work involved
- Built a self-updating research database integrated with EndNote, collecting studies and sorting them by country and disease automatically.
- Built a Python pipeline that processes thousands of research papers and summarises their empirical results into per-country slide decks, used to identify countries with comparable epidemiological profiles.
- Integrated the model with EndNote so the decks update themselves from a user prompt rather than being rebuilt by hand.
Stack
Python · NLP · Data pipelines · EndNote · Clinical research
Have something like this?
Describe the problem and you will get an honest answer about whether I am the right person for it. The scoping call costs nothing.