In Case You Missed It: Three Things we Learned in our AI Forecasting Webinar

If you missed our recent webinar, AI and ML in Pharma: Redefining the Forecasting Landscape, then I hope it’s because you were on a lovely vacation somewhere. Now that you’re back, though, you’ll be delighted to know that it’s available on demand so you can catch up at your leisure.
Our panel, which included Daniel Chancellor, VP Thought Leadership for our parent company, Norstella, as well as David James and Stefano Driussi both of J+D Forecasting, had a fascinating discussion about the potential applications of artificial intelligence and machine learning in pharma. AI is a transformative technology and it has the potential to radically change the future landscape of pharmaceuticals. Our focus for the webinar was around forecasting (of course!) and how forecasters can access some of the opportunities – and mitigate some of the risks – that AI can bring to the forecasting process. Here are three things we learned in the session.

1. 1. 1. Investment in AI is expanding beyond science. While improving clinical trials and accelerating drug discovery remain some of the key focuses of AI in pharma, there are businesses seeking to focus on the commercial aspects. For example, using AI to analyse vast amounts of information to help identify analogues. We’ve also seen organisations trying to set up models that look at things such as ROI of promotional spend, essentially using AI to be more effective at the commercialisation of drugs and how to spend money to maximise their return.
      2. People are still critical to the process. AI isn’t taking over just yet, in fact one of the biggest challenges companies face is finding people with the skills to run these very ambitious AI programs. Many of our clients have identified an agency or a partner to support them with the very specific skills that they require. The good news is that once they do that, pharma has also been very good at capturing and structuring data. All companies now have data lakes which they can tap into. There is huge opportunity in the commercial area which our panel felt remained relatively untapped.
      3. There is huge opportunity in unstructured data. There are several areas where forecasters can seek unstructured data to add to their models. The first is social media – this inevitably needs a lot of cleaning up, but there is great value to be found here. Another source, which is primarily for the US market is claims data that companies can use to inform models. Finally, there are physician notes which provide great insight into the ways that the medical professionals are talking about and treating diseases. All of these have great potential that can only truly be tapped with AI unless, as Stefano notes “you have several centuries” available!

   
   The webinar touched on much more than I’ve shared here so I recommend finding a few minutes to join the on-demand version. The team from J+D Forecasting has been part of Evaluate since the beginning of the year, and this is a great opportunity to learn about their expertise and to uncover potential uses of AI for your forecasting activities.

   If you’d like to check out our other on-demand webinars, you’ll find them on our webinar page.