Conversational AI watch: Ramco CHIA & the Self Explaining Payslip

This AI based service explains HR and payroll department queries while the chatbot is solving hybrid working model challenges.

As COVID-19 pushes hybrid working model into the limelight, 80% of businesses believe they will adopt the conversational UI approach in this decade. According to market reports, the global conversational AI market will be worth around US$ 16 billion by 2024. Organisations around the world are increasingly seeking ways to enhance employee experience, and digital transformation of payroll is one such avenue with immense scope.

Ramco Systems, a leading software provider, recently launched the Self Explaining Payslip (SEP), a smart, AI-based service that lets employees seek explanations of their compensation and payslip component.

Read more: Six reasons why HR is taking centre-stage in the post-COVID-19 workplace. Is your startup geared up?

Offered on Ramco’s virtual assistant chatbot, Ramco CHIA, the SEP is a move towards the Conversational UX approach, aimed to help employees receive prompt and personalised responses to their payroll queries, without having to write to a payroll helpdesk and wait for a response.

The Tech Panda spoke to Rohit Mathur, Head of Ramco Global Payroll and HR Solution, Ramco Systems, about their conversational AI systems and chatbots.

Given the current situation where the hybrid work model is becoming the norm, employees and even helpdesk agents might not be working from office every day. This can further exacerbate problems related to employee services such as access to resources or information, waiting time, and hence employee experience

“Given the current situation where the hybrid work model is becoming the norm, employees and even helpdesk agents might not be working from office every day. This can further exacerbate problems related to employee services such as access to resources or information, waiting time, and hence employee experience,” he says.

“With a digital, fully automated smart service such as the self-explaining payslip, it is like having access to a personal helpdesk agent, anytime, anywhere, day or night, 24×7, fully secured and protected.

The AI-based service can also help send proactive alerts and notifications to employees, for instance notifications about their tax investment declarations, thereby automating mundane tasks from the HR and payroll department,” he adds.

Features

Ramco leverages AI-based deep learning and NLP capabilities to improve HR and payroll process efficiencies and employee experience.

“Traditionally, HR applications provide dedicated employee and manager self-service portals, through which end users access such info. While these are still relevant, they can be cumbersome when the employee wants to do a quick transaction, like, apply for leave, check payslip or approve team member’s expense claims. Ramco’s AI-based digital assistant Ramco CHIA uses NLP to converse with employees and get such jobs done much faster,” Mathur explains.

Unlike the traditional self-service portals, which have to be accessed through dedicated URLs and menus, CHIA can be plugged into those natural digital workspaces where employees work day-in and day-out, for example Microsoft Teams.

Available on over 50+ messaging platforms, including Microsoft Teams, Slack, Workplace from Facebook, WeChat, Viber and WhatsApp, Ramco CHIA addresses a holistic set of 50+ employee and manager self-service functions related to leave, timesheet, travel, expenses, explanation of company policies and team management, amongst others.

Offered through CHIA, the SEP service helps employees receive prompt and personalised responses to their payroll queries, without having to write to a payroll helpdesk and wait for a response. Some examples of the kind of payroll enquiries that the SEP can help address include overtime/loss of pay break-up, arrears/back-dated payments, year-to-date and historical enquiries, month-to-month comparisons and tax computation worksheets.

In a simplified manner, Ramco’s CHIA has four parts to it:

1. An NLP engine that processes what an employee is querying for and provide a human-understandable answer. The NLP engine can understand several languages, enabling employees to use their native language for conversing.
2. Authentication mechanisms and protocols to validate an employee and ensure that the information is shared in a secure manner.
3. The underlying HR and payroll platform that stores all employee related details.
4. A log of both success and failures. Failures denote cases where the chatbot was not able to respond to the user in a satisfactory manner. This log is used to train the chatbot to get better.

How it Works

CHIA utilises AI in three ways to best serve its users. The first part is in the NLP engine that powers CHIA’s capabilities to understand what the users are asking.

“Users can communicate with CHIA as they would with a human being,” says Mathur.

Users can communicate with CHIA as they would with a human being

CHIA is equipped with linguistic comprehension capabilities enabling it to identify the language with which a user is communicating and respond back in the same language.

The second part is in Ramco’s Universal Personalisation Engine, a real-time personalisation engine that helps CHIA learn an employee’s personal preferences and choices for items such as leave type, durations, and reasons for leave and travel destinations.

The third part is behind the hoods, how does CHIA explain a salary component.

“This is powered by Ramco’s Payroll Reasoning Engine, which analyses employee salary data (what is normal vs. what is not), rule patterns (is the pay element affected by unpaid leave days?) and input triggers (did the employee get a salary hike in this period?) to determine how the pay elements of an employee’s payslip were calculated,” he explains.

Growth Story

Since its introduction a little over three years ago, CHIA has been used by over 40 customers, with a total of 1,50,000 users and Mathur says has helped save more than 26,00,000 minutes of their time.

A few of its customer profiles and examples include a manufacturing conglomerate with approximately 15,000 employees across more than 10 countries and a Middle East based staffing company with approximately 6,000 employees.

Read more: HR tech: ‘2021 will witness a massive digital transformation if not anything more’

To date, Ramco CHIA has clocked more than 1,20,00,000 conversations overall, across 30 channels such as WhatsApp, Facebook Messenger, Slack, Telegram, email, and Microsoft Teams.

The virtual assistant has answered close to 10 lakh queries raised by its users, such as requests to view payslips, leave balances, approval requests from subordinates as well as explanations for questions such as “Why are the deductions on my payslip high?” CHIA responds to a combined total of about 2,50,000 conversations every month.

“For an organisation with an employee count of 1000-1500, we typically observe 400-500 monthly logical transactions carried out through Ramco CHIA,” Mathur says proudly.

Future Plans

While digital payments and mobile-wallets are already popular in large Asia-Pacific countries such as India and China, the pandemic has further accelerated the adoption in other Asia-Pacific and European countries. The pandemic has also made employee wellbeing a key priority for organisations around the world.

By integrating such services into payroll platform, Ramco’s CHIA will be augmented to provide smart advisory services around employee wellness and financial intelligence

For example, Earned Wage Access (EWA) is becoming a key tool for employers to financially support their employees in a timely fashion. Keeping this progress in mind, Ramco has plans for the future work environment.

“By integrating such services into payroll platform, Ramco’s CHIA will be augmented to provide smart advisory services around employee wellness and financial intelligence. Thus, Ramco’s customers will be able to use CHIA as a common service channel for a variety of HR and Payroll services for their employees,” Mathur concludes.