IBM Watson Anywhere Updated to Overcome AI Roadblocks

Citing studies showing AI adoption is growing despite problems such as vendor lock-in, IBM announced new features for its Watson Anywhere approach designed to speed AI adoption and scale across any cloud and include multiple data sources.

Watson Anywhere was developed to help overcome obstacles to AI adoption including growing data complexity, issues with data preparation, and skills shortages, IBM said in an announcement on Monday.

The company cited several independent AI surveys, including Gartner's 2019 CIO Agenda survey showing that in the past two years AI deployments have grown from 4 percent to 14 percent. That is relatively modest growth considering the results of a 2019 MIT Sloan Management Review and Boston Consulting Group study, showing 9 out of 10 respondents agree that AI represents a business opportunity for their company.

IBM sees the gap between the enthusiasm for AI and its actual adoption is caused by complexity encountered where an organization is working with multiple cloud environments. It cites its own 2018 IBM Institute for Business Value study that found 76 percent of organizations reported using at least two to 15 hybrid clouds, as well as 98 percent planning to be on multiple hybrid clouds in three years.

Watson products running on IBM's Cloud Pak for Data will work across those cloud environments regardless of the vendor, the company said.

"We collaborate with clients every day and around the world on their data and AI challenges, and this year we tackled one of the big drawbacks to scaling AI throughout the enterprise – vendor lock-in," Rob Thomas, general manager, IBM Data and AI, said in a press release. "When we introduced the ability to run Watson on any cloud, we opened up AI for clients in ways never imagined. Today, we pushed that even further adding even more capabilities to our Watson products running on Cloud Pak for Data."

In support of the Watson Anywhere approach with Cloud Pak for Data, Big Blue included quotes from KPMG, and Air France-KLM on how they are using Watson apps to develop AI tools for their applications.

KPMG worked with IBM on AI apps that can use data from any source regardless of what cloud or AI platform is being used, the IBM announcement said. The resulting KPMG AI in Control solution leverages the Watson OpenScale platform to enable continuous evaluation of machine learning (ML), AI algorithms, models and data.

Working with Air France-KLM, IBM used Watson technology to create a customer assistance app called My Interactive Assistant (MIA), Big Blue's announcement said. The app uses IBM Watson Assistant with Voice Interaction to work with customers over the phone. MIA asks the customer for their reservation number and can pull up data including name, flight number and telephone number. If the callers need human assistance, MIA will pass that data on to an agent to streamline the process. Since July a pilot program using MIA in one country has handled 4,500 calls from people seeking help with their travel plans.

New Features for Watson Apps and Tools
This week's IBM announcement for the Watson Anywhere approach added new features and functionality for its products including:

  • Drift Detection for Watson OpenScale, which detects when and how far a model "drifts" from its original parameters and issues an alert. It is designed to improve the accuracy of models.
  • Watson Assistant, IBM's conversational AI product, now includes Watson Assistant for Voice Interaction that makes it easier for customers to ask questions in "natural language." It can pick up on nuances in the way humans speak to help find the appropriate answer. It also can combine texting with voice.
  • Watson Discovery search product, which uses ML and natural language processing, now includes Content Miner for searching large datasets for specific text and images.
  • Cloud Pak for Data, IBM's integrated data analytics platform, is now certified on OpenShift to assure that all the components come from supported sources, container images contain no known vulnerabilities, and are compatible across Red Hat Enterprise Linux environments in any  private, public or hybrid cloud.
  • OpenPages with Watson version 8.1, IBM's latest governance, risk and compliance (GRC) platform, has a new rules engine, as well as new intuitive views, visualizations, and workflow features.

For developers working on AI projects with the new technologies, IBM offers a new Cloud Pak for Data Developer Hub with tutorials, code patterns, and hands-on labs.