IBM has released IBM Watson AIOps to give the CIO superpowers. The CIO role is getting increasingly complicated and IBM says that AI can help reduce that complexity and lower stress. Watson AIOps will predict, self-diagnose and responds to IT anomalies in real-time. If effective, this means stopping IT outages that industry analyst firm Aberdeen claims can cost businesses US$260,000 per hour.
According to Rob Thomas, Senior Vice President of IBM Cloud and Data Platform: “The frontier of automation that really nobody’s taken on yet is using AI to automate many of the difficult things that we ask a CIO to do in her job every day. Whether that’s fix problems, prevent problems, predict problems, this is the kind of stuff that AI is really good at.
“We are pleased to unveil Watson AI Ops, which is the first AI system designed to give a CIO superpowers. To give them the ability to predict issues before they ever happen. If they happen to happen, fix them quickly, typically using AI, as opposed to having to call somebody and get them on the phone.”
How does Watson AIOps work
Watson AIOps uses analytics and machine learning to understand key business workflows and behaviour. It monitors and ingests data from multiple data sources such as log files and tools.
Watson AIOps alerts users to issues via a message in the collaboration app, Slack. It provides a report containing essential information about the issue, a pointer to where it is and a list of all affected services. It also provides the user with synthesised advice on how to diagnose and resolve the issue.
Jessica Rockwood, Vice President of Development, Hybrid Multicloud Management, IBM, explains how Watson creates the report. “Watson AIOps has gripped a diverse set of log anomalies and alerts based on spatial and temporal reasoning, as well as similarity to past situations. The algorithms connect my siloed data sources to synthesise a holistic problem report.”
The application allows the user to drill down into the data. Users can see why an event was included and see the key data points that were correlated. Rockwood says that this allows the user to: “Gain confidence in the analysis with the link to the originating tool.”
If this is a new issue to that user, Watson AIOps can also show a list of past incidents and rank them as to how similar they are. Rockwood notes that: “By using the NLP technology that IBM has added to Watson, it is able to understand the content in tickets to automatically identify and extract resolution actions.”
There is a caveat here. Watson relies on the completeness of the help desk tickets. Operations and help desk teams need to make sure that tickets are accurately completed. For a lot of organisations, the implementation of Watson AIOps is likely to highlight weaknesses in existing ticketing solutions.
Three additional features for the CIO
In the press release, IBM has highlighted three additional features to help the CIO. They are:
- Automate Business Planning – IBM Cloud Pak for Data, IBM’s fully-integrated data and AI platform, has been updated with a host of new capabilities designed to help business leaders automate the access to critical business-ready data.
- Automate Business Operations – A major new update to IBM Cloud Pak for Automation, software for designing, building and running automation apps, enables clients to more easily create AI “digital worker” automation solutions. The new capabilities can help simplify how organizations digitize automation skills, such as data capture, task automation and business routing.
- Automate Call Centers – IBM Watson Assistant, now has a pre-built user interface that requires no development effort to deploy and is designed with user experience-based best practices. There are also new integrations to some of the leading customer service platforms and a new feature called “autolearning”. The latter is currently in development and due in the product this summer.
Enterprise Times: What does this mean
Watson AIOps is just the latest member of the growing IBM Watson family. Rockwood says that there are over 100 patents in the product, all of which come from IBM Research.
Using AI to augment operations in the same way it is augmenting security teams makes sense. There is a lot of similarity in the way both teams use AI to identify and evaluate incidents. It will be interesting to see how quickly IBM gets this to customers. As companies move forward, they will be looking for any advantage they can get.