Dharmendra Sahay co-wrote this blog post with Arun Shastri.
This post is the first in a two-part series.
In a few days, we’ll be presenting our thoughts on how to create impact with artificial intelligence at the 2018 PMSA Annual Conference. We’ll talk about how AI is being used in life sciences and how AI could be used, and we’ll bust some myths. We’ll also share detailed advice on how to start or expand an AI capability at your organization, which includes these five critical steps:
- Educate your stakeholders. Don’t assume that they understand AI just because they’re enamored by it. Executives need to understand that AI isn’t just about cool technology. It opens new possibilities for automation that can transformyour company’s business model. But AI applications in an enterprise may be very different from what people hear about in the media, such as image recognition on Facebook or voice assistant interactions through Alexa. Make sure that your executives understand what AI is (and is not) and what it takes to realize its potential.
- Think outside of your immediate scope. Don’t solve problems that are already solved well with traditional methods. You have to think about answering new questions, and leveraging new data sources that may present opportunities to solve new problems or tackle the same problems but in entirely different ways. Leveraging AI to squeeze marginal gains in effectiveness in solving current problems will not inspire confidence in the power of AI.
- There is no prescribed sequence in tackling problems. Create a balanced portfolio of adjacent, leapfrog and disruptive ideas. There is no rule that says that you need to start your AI journey with a full-scale, organization-wide diagnostic. You need your stakeholders to be excited, so ensure that high-impact projects are in your mix. Push your team’s boundaries and inspire them to pursue projects that grab attention around the organization early, so people say: “Wow, that’s clever! How can I get in on it?” That infectious enthusiasm can become the driver for a more methodical transformation.
- Don’t obsess over quick wins. Visible change will take time, as it requires a change in organizational mindset. For example, if you’re planning to leverage AI to drive suggestions on the “next best action” to sales representatives, finding the right algorithm and tool may be just one part of it. Driving the right experience for sales reps and developing early champions before scaling are equally important. All of this takes time and patience.
- Recognize that it takes more than data scientists to build an AI capability. Recruit people with the right talent and experience. If you’re going to leverage existing talent, make sure that the bar is set high for them to master new skills. To succeed, you need a team of more than a few advanced data science professionals. You’ll need data engineers, AI/ML software engineers, liaisons to the business, data scientists and more.
We’ll be sharing some more content from our presentation with you in an upcoming post. We hope to see you at PMSA and look forward to sharing more about creating impact with AI.
EVENT: PMSA 2018 Annual Conference
BLOG POST: Five Myths About Artificial Intelligence