CrushBank Brings Watson AI to the MSP Help Desk

By Frank J. Ohlhorst

Most any MSP understands the importance of the help desk. After all, a responsive help desk can make the difference between keeping customers long term or quickly losing them to competitors; yet, offering those help-desk services can be a real drain on resources and can increase operational costs significantly. What’s more, MSPs are finding that staffing help desks with qualified individuals, who also have a personal relationship with the customer, to be an almost impossible task.

With those concerns in mind, Evan Leonard and David Tan set out to come up with a solution to the help desk dilemma and founded CrushBank, a cloud-based intelligent help desk that leverages natural language processing and machine learning to create what amounts to a cognitive-computing-powered help desk.

Since its launch in 2015, CrushBank has come to encompass the primary functions of a help desk and has the potential to be a valuable tool for anyone operating one. Of course, IBM’s Watson is a big part of the CrushBank solution, however there is a lot more to CrushBank than a jeopardy playing smart machine.

The company took the approach of building a help desk from the knowledge-base point of view; in other words, what could be accomplished if your help-desk system was able to ingest all of the technical information surrounding IT systems in use at a particular site and then apply advanced search capabilities to that warehouse of data. Incorporating institutional knowledge was only the beginning; CrushBank also licensed the content from a vast array of technical manuals and books, as well as educational and support-orientated data.

Yet, having a plethora of technical information does not transform a knowledge base into a cognitive help desk — that is where Watson comes in. Watson is able to parse the data using natural language processing and develop logical relevance to any queries inputted into CrushBank. In other words, help-desk operators simply type a query into the system, such as “TCP/IP Error code 110 on connection to firewall” and the system will search out all relevant information, including any other occurrences of the error, as well as possible solutions and present that information to the technician in a matter of seconds.

SInce the system also stores historical data and can ingest the data from previous help-desk and support systems, everything relevant to the particular location or network will be readily available, as well as any previous instances of that particular error, or any errors related to the situation. With Watson behind the scenes, problem resolution improves the more the system is used.

There are numerous advantages offered by the pairing of Watson with troubleshooting data, the first of which is an ever growing depository of problems and solutions, which can be leveraged quickly and easily to bring new help desk staffers up to speed. Other advantages include the ability to ingest new technical information as new devices or services are added to IT, while also incorporating policy and other types of intellectual property into the system.

CrushBank has the potential to help MSPs reduce training costs, while also addressing the problem of employee turnover. By making it easier for staffers to be successful in their jobs, frustrations can be reduced. What’s more, the system records …

… queries and interactions, meaning that knowledge is preserved, even if a help-desk technician moves on to a different position at a later date.

As a cloud-based system, CrushBank proves rather simple to deploy and the company offers API integration, allowing it to be integrated with existing systems, so the investments in those systems are preserved.

CrushBank also is an example of the possibilities present by integrating cloud services with AI/ML services such as IBM’s Watson and proves to be a good example of how normally disassociated data can be combined into useful information. While CrushBank might not solve every MSP’s help-desk challenges, it proves to be a good indicator of what can be accomplished when you look at solutions to problems from a different point of view.