HIIT 2.0: The Incremental Benefits of Machine Learning

[dangrady510]

Like many of us, I could benefit from losing a few pounds (maybe more than just a few).

So, as part of my most recent unrealistic attempt to get back to a weight that our Wii Fit doesn't consider obese, my wife and I went out and got a Peloton bike.  On a recent ride, one of the instructors said "You body isn't Amazon Prime, it isn't going to show up in 2 days." This got me thinking about something other than how exhausted I was.

I started thinking about all the improvement initiatives in life that we start and then ultimately give up on too quickly because we don't see dramatic results immediately or in the short term.  My mind wandered to focus specifically on improvement initiatives in the workplace because I work at ServiceNow and making the world of work, work better for people is so ingrained in my DNA now I even think about it when I'm pedaling along at 18 miles an hour up an imaginary hill.

For a number of years now business leaders everywhere have recognized the potential of Artificial Intelligence - and its subset machine learning - to harness data to accelerate and drive digital transformation.  However, many have fallen victim to Amara's Law which states: "We tend to overestimate the effect of technology in the short run and underestimate the effect in the long run."

Out of the gate, many organizations struggle with AI and machine learning in part because their vision of AI's impact is too big and they can't define practical applications that would generate short term value.  Some struggle because they don't have the data science skills or infrastructure necessary to act on impactful applications of machine learning.  Ultimately, there struggles can lead to frustration, lost momentum, and bring opportunities for improvement to a stand-still.  Some even give up entirely.  Which doesn't sound too bad right now on the 3rd minute of this incline I'm struggling with.

This is where ServiceNow's in-platform machine learning - branded Predictive Intelligence - comes in.

Predictive Intelligence provides a layer of artificial intelligence that empowers features and capabilities across the ServiceNow solutions to provide better work experiences.

This AI layer delivers a set of machine learning frameworks that support focused solutions to common challenges organizations face, but can also be used as the foundation for innovative solutions beyond the ones we provide.   Let's take a look at the different frameworks available today:

Classification Framework - enables you to use machine learning algorithms to set field values during record creation, such as setting the incident category based on the short description.  You can train predictive models so they act as an agent to automatically categorize and route work based on past handling experience.  Applying the classification framework in this way allows organizations to handle higher volumes of incoming requests at lower costs.  Error rates and resolution times go down, satisfaction scores go up.

Similarity Framework - uses machine learning to identify existing records that have similar values to the record someone is currently working with.  This goes beyond just simple string matching to help surface more relevant content.  For example, you can train a subset of your incident records to recommend a resolution based on the information of a similar incident record.  By reusing similar closed incidents that have a proven resolution, you can eliminate a lot of the hunting and pecking that goes on in most organizations and help agents more quickly provide the best resolution for incoming incidents.  This framework can also be used to surface more relevant Knowledge articles or alert someone creating a Knowledge article that a similar article already exists, increasing overall productivity in the organization. 

Clustering Framework - can be used as a discovery tool to automatically group similar records into clusters so you can address them collectively or identify patterns.  For example, you can quickly identify patterns in recent incidents to identify where a major incident should be created.  Another scenario would be to group similar search terms together to identify potential future Knowledge articles.

Natural Language Understanding(NLU) - is an inference service that you can use to enable the system to learn and respond to human-expressed intent. You train the model based on the vocabulary of your own organization to help it understand word meaning and contexts so it can infer the action that an employee or customer wants to take.  This framework is initially focused on improving the experience that employees and customers have with the ServiceNow Virtual Agent.

This library of frameworks and targeted solutions continues to grow with every release of the Now Platform.

By leveraging an in-platform AI engine like Predictive Intelligence, these modern solutions become more accessible.  By lowering the barrier to entry and providing more focus with some targeted solutions organizations can make the most of limited resources and start to reduce manual intervention, improve customer satisfaction, and elevate employee productivity.

Much like my weight loss initiative, we may not see the results in two days, but the incremental improvements will be obvious and encouraging.  That encouragement will be enough to get you are your organization over the very real hill of deploying practical AI solutions.

I just hope the Peloton instructors little words of encouragement are enough to keep me pedaling up these imaginary hills they keep throwing in front of me.

As long as we keep moving we'll both end up where we want to be.

Although with 14 more minutes to go on this ride, I'm looking forward to the day where I can just order some "weight loss" on Amazon Prime and have it show up in 2 days.

Other Blogs Related to this Topic

Hidden Now Intelligence Gems of the Paris Release

Knowledge Demand Insights: How to get Smarter about Getting Smarter

Natural Language Understanding(NLU) Makes the Virtual Agent the Road More Traveled

The Digital Ensemble that Helps Create Great Virtual Agent Experiences

Predictive Intelligence and Virtual Agent - Better Together

Diversity, Inclusion, Belonging: A Data and Analytics Perspective