vasabii - Fotolia

IBM cognitive computing: Tips for starting a channel practice

At IBM's latest partner conference, platinum-level IBM Business Partners Perficient and Mark III shared the routes they took to build cognitive computing practices.

Channel firms looking to build cognitive computing practices can choose from multiple paths.

The IBM Business Partner Leadership Conference, held this month in Las Vegas, showcased partners' variety of options for getting started and asserted IBM cognitive computing is accessible to all. The company also argued that cognitive capabilities are rapidly becoming vital for differentiation -- "the essence of competitive advantage," IBM chairman, president and CEO Ginni Rometty said. Other IBM executives urged partners to recognize opportunities in the emerging decision-support market, reachable only to partners that incorporate machine learning, artificial intelligence (AI) and cognitive technology into their offerings.

Perficient applies IBM cognitive computing to healthcare

John Kelly, senior vice president and director at IBM Research, said partners could start IBM cognitive computing practices by delivering products through the cloud, building their own, reselling IBM's portfolio, or enhancing and expanding the existing technology, among other avenues.

"It's pretty clear that sometimes people say, 'Well, AI, Watson -- this gets very complicated.' No. There are many, many ways to get started that are very simple to produce lots of capability," Kelly said.

Numerous IBM Business Partners illustrated Kelly's point, providing conference attendees examples of how they have used the Watson platform in an array of industries.

If you are not planning a [cognitive computing practice] already, you are too late.
Michael Gerentineglobal vice president of business partners and marketing, IBM

Perficient, a St. Louis-based consulting firm and Platinum IBM Business Partner, worked closely with IBM to develop its Watson practice before setting off on its own. The company today has about 22 customer engagements, said Kevin Nunnally, national partner executive for IBM sales at Perficient.

Nunnally told attendees about using Watson in the healthcare industry, a segment he said represents about 30% of Perficient's business. "Moving into the cognitive aspect of healthcare was something that was very important to us. We saw a great opportunity with the skills we already had in healthcare," he said.

One of Perficient's clients, TriHealth, a large healthcare provider based in Ohio, struggled with predicting patient readmissions. "If you're a hospital, [patient readmission] is an important problem to attack ... from a patient care standpoint and a financial standpoint." The healthcare client could predict about 46% of its readmission cases, Nunnally said, falling short of Medicare's minimum standard of a 70% predictive rate required for reimbursement.

First, TriHealth implemented IBM SPSS, a predictive modeling tool, which allowed the organization to use its structured data to increase its predictive rate to 71%, Nunnally said. Working with Perficient to further improve the rate, TriHealth realized about 20% of all their available data is structured, the rest of it being unstructured. "That unstructured data comes from physician notes [and] other healthcare provider input that goes into an EMR [electronic medical record]," he explained.

To get that unstructured data, Perficient used the Watson Explorer content analytics tool paired with IBM's Healthcare Annotator. "Using the Healthcare Annotator, they were able to analyze that other 80%" of unstructured data, structure it and feed it to the SPSS predictive model, he said. "Through that process, [TriHealth was] able to increase the prediction on readmission from 46% to 71% to 93%."

Mark  III develops the BlueChasm unit

Mark III, an IT solution provider based in Houston and Platinum IBM Business Partner, has mirrored its evolution with IBM's throughout its 21-year-old relationship, said Andy Lin, Mark III's vice president of strategy.

The company, founded with a core focus on IT infrastructure and the data center, which still comprises a large portion of its business, has more recently adapted to IBM's digital transformation strategy with Watson and Bluemix, Lin said. In late 2014, Mark III spun up a digital development unit called BlueChasm that it headquartered in Austin, Texas, a city he noted that is generally known for its developer talent. The BlueChasm unit develops open cognitive platforms on the IBM Cloud in part by using Watson and Bluemix APIs.

"What we do [at BlueChasm] is basically try to build platforms based on what we see in the market and what our clients tell us they need," he said. "We don't go out and try to guess. We build it directly on client feedback."

BlueChasm has since built a cognitive call center and a video analytics offering. While the BlueChasm unit "operates almost separately" from Mark III, he said, the company has discovered synchronicity between the BlueChasm and Mark III practices.

"We have come to evolve into what I call 'the IBM full-stack partner' where we can bring in those cognitive capabilities, the digital capabilities ... but if [clients] want to refresh their ERP application, modernize what they are doing there ... we can absolutely do that as well."

"I think that is a little unique in the channel ecosystem. I know the amount of effort it took to get us where we are, and it takes unbelievable people," he added.

In Lin's opinion, the first step for channel firms that want to start a cognitive practice is to take stock of their own teams. "Understand what your team does best and what their skills are and maybe what their hidden skills are." He also recommended channel firms look at their current customer base and customers' industry focuses and specializations.

"Build your strategy around those things, around the strengths of your team. ... You would think it's common sense -- look at what you have and then base your plan around it -- but I think you'd be surprised how many times it doesn't happen."

In terms of hiring talent, he said that talent is out there, but sometimes exists "outside of the ecosystem."

"Sometimes you have to be willing to take a chance on people and evaluate people based on raw skills mapped onto what your objectives are, and then give them the space to do what they do best," he said.

Other Watson developments

IBM has established resources to jumpstart partners' cognitive businesses, noted Michael Gerentine, global vice president of business partners and marketing at IBM. Among those resources is Watson Build, a contest encouraging partners to develop Watson applications. IBM will award the best partner-built application at the next World of Watson conference this fall.

In addition, IBM and distributor Avnet recently revealed an alliance to speed up the time of market for IoT products. Through the partnership, the companies will collaborate to fast-track Watson IoT and Bluemix solution development. Avnet will make more than 150 IBM Bluemix cloud services available through the Avnet Cloud Marketplace. Partners will also have access to Avnet's training and education services to gain Watson IoT and Bluemix expertise.

IBM cognitive computing: Partners should move quickly

IBM, while highlighting the accessibility of IBM cognitive computing, goaded partners to start practices right away.

Michael Gerentine, global vice president of business partners and marketing at IBM, told SearchITChannel that partners in or transitioning to the cloud world will find cognitive opportunities adjacent to those efforts.

"In order to build a cognitive solution, it should be cloud-based, which means you should be able to build on the IBM Cloud, which is a SaaS reoccurring revenue model," he said. "By default, by building a cognitive solution, you are moving to a reoccurring revenue model, and that is how you are going to charge your client for that solution."

He warned partners not to lag behind. "If you are not planning a [cognitive practice] already, you are too late," he said.

Next Steps

Learn more about machine learning opportunities

Read about the natural language process marketing

Can IBM Watson defeat today's cybersecurity threats?

Dig Deeper on Emerging Technology Markets