We are now part of Eviden, discover more...

CAO Fall Recap Take 2: How To Successfully Implement Machine Learning

In a previous post, we shared the recurring question from the first day of the Chief Analytics Officer fall conference: how can you best apply insights from your data? In response to that question, several presentations offered guidelines for building machine learning models and shared case studies showing how companies are currently utilizing these solutions.

During the remainder of the conference, while many speakers focused on different aspects of machine learning and their specific solutions, one thread linked all of the presentations together: solutions must continually adapt over time. As Horia Tipi, Vice President and Head of Global Optimization of FICO said, “Adaptive analytics is the notion that the solution you deployed is merely version one, and tomorrow you may need version two.”

In a session specifically focused on how to make machine learning work for your organization, Varun Manocha and Vikas Sharma, both Vice Presidents, Decision Analytics at EXL, described the challenges enterprises face and the considerations when implementing machine learning. It’s no surprise that several factors need to work together to contribute to the overall success of a machine learning program, including talent, big data infrastructure and data accessibility.

They used an example of the development of the modern automobile as an analogy to demonstrate the three basic stages needed to build a machine learning platform for your business:

  • Pilot: The initial step in the creation. For a car, this was the development of the first self-propelled vehicle. In terms of machine learning, this would be initially identifying use-cases with high value and implementing simple algorithms for the proof-of-concept.
  • Early Use Case: For the development of the car, the early use case was the first vehicle to be propelled by an engine. For machine learning, this is an extension of the success of the pilot through the exploration of complex algorithms and understanding the dynamics of scaling.
  • Industrialization: The third phase in the evolution is bringing it up to scale. For cars, this phase brought about mass production. For a machine learning solution, this means leveraging big data and advanced algorithms to establish governance for the machine learning enterprise.

By following these three steps, enterprises can build more effective machine learning solutions to improve their operations. Varun suggested starting with a model with a low cost-of-error for your company, such as a personalized recommendation system. “Even if you recommend (the) wrong product to a customer, things will not break,” he noted.

Another presentation took this concept one step further and detailed a real-world application of machine learning from idea to implementation. The speakers explained how Mount Sinai Hospital in New York is using machine learning and natural language processing (NLP) to improve the healthcare experience of its patients. Poornima Ramaswamy, Vice President, AI & Analytics Practice at Cognizant – the company to deploy and manage Mount Sinai’s machine learning solution – kicked off the session. Ramaswamy explained that personalization and transparency are of utmost importance to gain client and user trust when implementing AI systems. If a company does not prioritize these when building their solution, she said, AI risks failure.

Ramaswamy then introduced Varun Gupta, IT Director of Advanced Analytics and Data Management at Mount Sinai. He described how each of their patients, on average, will interact with 15 to 20 separate systems across the span of their care. As such, Mount Sinai faced the challenge of building an ecosystem on top of all of these data-intensive platforms to collect and mine the pertinent information. Their engineers built a hybrid cloud solution to gather the data, and then they used NLP to analyze it. This enabled the hospital to humanize that data and break the patients down into different buckets to make personalized recommendations for their care.

Could your organization benefit from implementing a machine learning solution? As one of just a few companies worldwide that has achieved Google’s Machine Learning Partner Specialization, Maven Wave is well-equipped to help companies develop innovative solutions rooted in machine learning. Contact us to get started!

About the Author

Kylie McKee
Kylie McKee is a Content Marketing Strategist at Maven Wave with more than eight years of tech industry experience and five years of content marketing experience. Prior to joining the Maven Wave team, Kylie worked as a Content Marketing Specialist for WebPT, Inc. and earned an Associate in Applied Science in Motion Picture, Television, and New Media Production with a CCL in Screenwriting from Scottsdale Community College.
October 17th, 2018

Get the latest industry news and insights delivered straight to your inbox.

Sign up for our Newsletter