New Infosys AI tool could transform the way companies maintain complex systems

Like so many organizations today, Infosys, the Indian consulting giant, is a company in transition. For years it has made a good living helping customers manage legacy tools, but CEO Vishal Sikka, who took over 21 months ago saw a shifting landscape and he began implementing new systems immediately.

One of those changes involved developing a new artificial intelligence system they have called Mana, which is designed to help customers automate repetitive system maintenance tasks and build knowledge about the underlying systems using artificial intelligence and machine learning. Sikka announced the system this morning at the Infosys Confluence conference in San Francisco.

Mana involves three main tools: Infosys Information Platform for analytics, Infosys Automation Platform for automating and continuously building knowledge about system maintenance and workflow tasks and Infosys Knowledge Platform, a formal platform for capturing and storing knowledge.

Infosys understands first-hand the huge amount of work and energy that goes into maintaining large legacy systems, whether computers or mechanical systems like turbines, locomotives and airplane engines. The company has more than 11,000 people alone working on fixing software bugs, a time-consuming and tedious task, which involves not only identifying a bug and a fix, but also the dependencies that fix could affect (and likely break).

In the end, these tasks can take weeks of employee time and wads of cash to fix and they are ongoing and virtually endless in legacy systems. With Mana, Infosys customers can begin automating large chunks of the tasks associated with this type of maintenance.

“You can now do simple maintenance in an automated way. This has a radical impact on usage and lifecycle management of these systems,” Sikka told TechCrunch.

At the same time, there is a growing knowledge gap as older employees, who have spent years maintaining these systems begin to retire and take their knowledge with them. In the 90s, companies tried to capture this knowledge in Knowledge Management Systems, but the systems were large and complex and getting people to enter what they knew (even if they could articulate it) proved challenging.

Enterprise 2.0 tools like Yammer and Jive were supposed to change that. After all, people shared their knowledge simply through the act of chatting in the social tool. All you needed to do was search for it (if you knew what you were searching for). It was a bit better than the monolithic knowledge systems, but it didn’t really solve the problem.

Infosys thinks artificial intelligence and machine learning technologies developed more recently and packaged in Mana could help solve knowledge management and tedium of system maintenance once and for all.

Infosys believes that once you begin to get a grip on these basic plumbing issues, you can put humans to work on more fundamental problems that require creative thinking, something machines are a long way from achieving. A more cynical view is that companies will lay off the employees once they are no longer needed. The reality is probably going to be somewhere in the middle.

Sikka says that every new efficiency brings new jobs and new problems to solve, problems that still require human ingenuity, and that’s not going to end any time soon. Mana — and by extension machine learning and artificial intelligence — is just another step in the evolution of technology and automation.