A global top 10 pharma client sought our support to understand how Generative AI platforms like ChatGPT can be used to generate plain language summaries (PLS) from scientific publications. They also wanted to understand the accuracy rate and production time.
We worked with the client to identify 1-2 scientific publications and plain language summary goals (desired output). Our team trained the generative AI model to generate a plain language summary, which was compared to the desired output. We documented processes, successes, and learnings.
Generative AI produced output showed significant accuracy relative to desired output. We explored potential automation areas for application of Gen AI across scientific and medical communications use cases.