AI in the Pharmaceutical Industry: Getting Drugs to Market Faster
According to Accenture, AI has the potential to save the US healthcare economy $150 billion in annual savings by 2026. With developments in IoT, voice recognition software, and a wide variety of AI solutions, it’s clear that AI in the pharmaceutical industry will be transformative.
To better understand how companies are embracing AI in the pharmaceutical industry, we turned to our customer Sanofi, a leading global healthcare company that strives to provide innovative healthcare solutions and improve access to treatments.
The Problem: Overcoming Bottlenecks with Drug-Discovery and Development
One of the biggest challenges facing healthcare today is quickly identifying, testing, and developing new drugs and getting them to the market as quickly as possible. Before patients can have access to a particular new medicine, it has to pass a very rigorous testing process that must be documented and submitted to the appropriate governing body, which has its own ever-changing rules and regulations. Once that’s complete, the drug can then go into production and can be sold to market.
When Sanofi was looking at ways to streamline this process, they identified bottlenecks within the data preparation, submission and report writing processes. As with many companies across different sectors, report writing relies on someone, usually a specialized analyst, looking at the data, summarizing it and explaining the insights in plain language. The process in the pharmaceutical industry is even more complex because these reports must also follow complex regulations.
Conventional Data-to-Insight Workflow:
The Solution: Automate Report Writing with Natural Language Generation
In order to help with this problem, Sanofi searched for an AI solution that could help address bottlenecks in their document preparation process and, most importantly, could be a technology that strives for 100% accuracy.
Machine Learning programs take a statistical approach to data analysis. The software makes decisions based on relationships that are learned from the data. While this can be a great solution in some use cases where there is a large amount of data and a tolerance for error, this was not the case for Sanofi.
Instead, Sanofi turned to Natural Language Generation (NLG): an AI software that reasons and provides a written description of what’s going on in the data. Sanofi chose to go with Yseop Compose, the only NLG software that can automatically generate reports describing what occurred, why, and recommend next steps, based on up-to-date best practices and relevant regulations.
As an initial project, Sanofi automated portions of the safety report, a document required by health authorities. Within one month’s time, the software was up and running, writing based on data-driven insights. With Yseop, Sanofi employees no longer needed to generate each of these sections from scratch. Now, they would just review the automated text, make adjustments if necessary, and then move on to their next task. Due to the success of the project, they’re set to automate more sections of safety report and are looking to automate reports from other departments.
On why Sanofi Chose NLG:
“We wanted to be certain and very accurate around what we produce in these documents that are submitted to health authorities. We had to rely 100% of what would be drafted.”
– Laetitia Marossero, Domain Lead Artificial Intelligence and Wearables at Sanofi
The Benefits of AI in the Pharmaceutical Industry: Saving Time, Saving Money, Saving Lives
- Saving Time: Generally speaking, this whole process of writing the report would take around 7 months from start to finish. The final result: Sanofi could reduce processing times up to 2 months, a time savings of almost 30%!
- Saving Money: Even after a month, Sanofi was getting a strong ROI. Employees tasked with writing this section of the report now could use their time more effectively and focus on more high-value, scientific sections of the report.
- Savings Lives: Sanofi focuses on many therapeutic areas (TA’s) like oncology, diabetes, cardiovascular diseases, vaccines, and more. By combining the benefits of saving time and money, drugs for these diseases can get to market faster at less cost.
Sanofi’s use case is just one example of how Natural Language Generation is being used today. There are many examples in Finance, Retail, Utilities, and much more. Get more details about popular NLG use cases and how to get the biggest ROI on your investment in our whitepaper.