NLP-based UVO not just responds to basic queries but also recognises subjective differences in how people tend to ask questions.
Robotics and AI are redefining various processes at the workplace across geographies; riding the wave, Tech Mahindra has developed some interesting technology in-house. In a bid to leverage advancements in digital technology to enhance employee experience, the company has developed a chatbot called UVO.
HR is always flooded by queries from employees regarding various policies and processes, taking a lot of valuable time away from other schedules. Such queries are traditionally directed to HR either through mails or meetings, which take up a lot of time and multiple transactions. At the same time, the existing ticketing systems for raising queries and requests also involve multiple approvals and are consequently more time consuming.
Recognising these challenges and sharing the purpose behind UVO, Kiran VS, head HR transformation, Tech Mahindra, says, “The idea was to move from the existing transactional interface based technology to a communication based interface.”
UVO (Ur Virtual Office Assistant) is also designed to automate resolution of employee query and with a broader intent to make it Conversation-as-a-Framework or CAAF, which can as well be built as an operating system.
“Looking at the challenges HR managers face and with two purposes in mind – easing the query resolution process and leaving HR professionals with more time, taking the redundant transactional tasks away from their plate, we looked at designing a consultative system that can support as a workplace buddy,” Kiran shares. UVO works as a virtual personal (digital) assistant to employees providing quick and convenient solutions to their day-to-day queries.
Unlike many other existing chatbots, UVO is NLP based and hence capable of not just responding to basic queries but recognising subjective differences in how people tend to ask the same question in different ways.
To enable this functionality in UVO in its first phase, Tech Mahindra identified the queries that went unsolved and found out that it was because either the bot didn’t understand the intent of the query or lacked an answer to the same. While the latter could be taken care of by correcting or adding to the menu of responses, the first challenge was where its NLP capability came into play to help comprehend speech and emotions for providing the right solutions.
To enable this functionality in UVO in its first phase, Tech Mahindra identified the queries that went unsolved and found out that it was because either the bot didn’t understand the intent of the query or lacked an answer to the same.
While the latter could be taken care of by correcting or adding to the menu of responses, the first challenge was where its NLP capability came into play to help comprehend speech and emotions for providing the right solutions.
UVO can also attend to individual specific queries such as sharing an employee’s leave entitlement or leave balance. Furthermore, it can take workforce actions such as applying for a leave, getting it approved and placing travel requests on behalf of an employee. The bot also offers 20-30 self-help videos offering basic information on policies and processes and helping employees utilise them optimally and conveniently.
“The bot eases the process of query resolution by simplifying information for the users and saves time and effort by providing instant and comprehensive responses,” says Kiran. It also features a reporting system that gathers insights from the kinds of queries raised and analyses employee concerns.
“Looking beyond, we are also developing the bot further such that it can ask questions seeking employee feedback and provide the requisite feedback to the concerned departments or managers,” he adds.
While the staff at Tech Mahindra enjoys the convenience offered by UVO, the company is further looking to enhance its offerings by developing sentiment analysis capabilities in the bot. With this in place, UVO will be able to suggest customised and individual specific wellness programmes to employees, taking the entire responsibility of their health and well-being. While the bot is already supporting employees at Tech Mahindra, this functionality is expected to be rolled-out by October this year.