In 2017, Apple’s iPhone X was a revolution of sorts. The smartphone came bundled with facial recognition (FR). This was the first time a mass consumer brand put such sophisticated technology in the hands of its users. To unlock the phone, all one had to do was show his or her face to the phone’s camera.
Ever imagined that one day, your face will be the entry pass to your office? FR is already a reality at Tech Mahindra. To stay cutting edge in the truest sense, the IT firm developed an in-house system FR system and incorporated it into its work-spaces about six months ago.
All the employee has to do is show up at the door; the facial scanning is almost instantaneous. It’s an efficient way of marking your attendance. Ask Harshvendra Soin, the man at the helm of Human Resource (HR) Management at Tech Mahindra. He says there is more to it than meets the eye. It’s not just attendance. It goes beyond that.
“Behind it is our Artificial Intelligence (AI) technology that recognises facial emotions. It captures eight emotions — happiness, anger, surprise, disgust, fear, contempt, sadness, and neutrality — every single day, for each employee that shows up,” says Soin, the company’s chief people officer. The ‘mood-o-meter’ consolidates the data into scores and sends it to Soin’s dashboard in the form of a ‘happiness quotient’ in Tech Mahindra speak. The scores range between one and five (with five being the happiest).
Soin monitors this data daily. “If it was 4.2 yesterday and today it falls to 3.6, it is an indicator for me to act,” he says. His team investigates whether the ratings are going up or down for an employee. The team also investigates in case it goes up, to know what they are doing right.
The company takes consent from its employees before registering them into the system. Once enrolled, they can opt out too. Interestingly, 95% of the employees have consented to this system. After all, who wouldn’t want someone to care about how they feel at work? The company claims the data is encrypted and deleted after processing.
Such a technology may be the stuff of science fiction for old-school HR managers, but it is real. AI and analytics-based tools are being used for people management in some organisations. Among these, IT firms such as Tech Mahindra and HCL Technologies are ahead of the curve in India.
Many HR tech start-ups have emerged in India over the past two years that provide services and products to companies across different sectors in India. HR tech is estimated to be a $500 million industry here. Globally, it is a $40 billion market that will grow exponentially in the coming five to seven years.
Technology has been part of HR management for over two decades. It began with Systems, Applications, Products or SAP-based systems in the early 2000s. More recently, however, advanced technologies such as AI, machine learning (ML), cloud computing, blockchain and so on have found application in HR management. In many countries, firms are now moving towards development of ML-based models, and have even delved into predictive analytics in HR. Using such technology, you can predict parameters such as attrition rate, recruitment needs and such.
“Earlier, technology was about doing day-to-day tasks, but now the HR industry is acting as a change agent and catalyst to make technology a way of life in the workplace — their new focus is around finding the right people, connecting people, engaging people, skilling people and replacing people — finding an answer about what to do when replacing people becomes imminent,” says Rituparna Chakraborty, co-founder and executive vice-president of TeamLease Services. Crucially, the decision-making process will be based on accurate historical data. This should effectively rid the system of subjective biases.
Facebook, Twitter and a host of other social networking sites have transformed the way we interact with people. We are hooked on to our apps and sites, and even businesses leave no stone unturned to chase us on these platforms. HR tech firms emulate such networks internally. Some of these systems have interfaces similar to Facebook and LinkedIn.
More than four years ago, HCL Technologies developed one such platform for its employees called Social. Anand R, senior vice president - HR, HCL Technologies, reasons why it was important for them to develop such a platform. “Today I need information to make HR decisions. A lot of valuable information often doesn’t reside in the formal MIS (Management Information System) or with the management,” he says. It’s with the people in the organisation, and it’s easy to keep track of people on platforms such as Social.
“As a HR manager, it helps you. For instance, I can know who is the most followed system architect on the app. I can identify a person whose ideas have repeatedly become high-value propositions for the company. He/she is a valuable asset for the company. Such people are twice more likely to be promoted than those who are not,” says Anand. The monitoring takes place only after taking the consent of the employee, of course.
Apart from having a lot of HR value, a social network also works as a ‘safety valve’ for the organisation. It is akin to having freedom of expression in a democracy. Adarsh Mishra, chief human resources officer (CHRO), Panasonic India, feels it is important for an organisation’s stability. Six months ago, the Japanese corporation launched its own internal social network and digitised several other aspects of HR.
“The average age of the workforce at Panasonic is 29 years. With that as context, we promote an open culture. People are encouraged to raise issues — cab or transportation issues, training issues, compliments, and complaints — over this platform,” says Mishra. It also works as an integrator for a workforce spread across locations, verticals, and divisions. As a result, it wants to use technology to keep them in the company for the longest time.
Hiring to retiring
Darwinbox is a three-year-old HR tech start-up headquartered in Hyderabad. It automates the entire life cycle of employees, from recruitment and performance appraisal to retirement. Some of its clients include Dr Reddy’s Laboratories, Wipro, Marico, and Delhivery and it has 300,000 employees on its platform.
Chaitanya Peddi, co-founder and product head, Darwinbox, feels that automation is cutting down time needed for several unimportant HR functions such as an employee help desk and CV shortlisting. “In a lot of companies, virtual chatbots are a reality now. You need not have a help desk for employees. You can ask chatbots questions such as what is the minimum period for a sabbatical or what are the criteria to apply for paternity leave.”
Using Darwinbox, client companies are even automating their recruitment process. “CV shortlisting is a tedious exercise. With machines, it is so easy. The job description clearly shows the level of experience we want. We also tell the machine who were the five people hired for the same role. It reads the CV’s personal and professional details, and can filter the top hundred guys out of 10,000 CVs in a matter of minutes,” says Peddi. Manually, the process takes several days. However, the automated route is taken only for junior and mid-level recruitment.
Software giant HCL Technologies is also applying AI-based solutions in its own recruitment process. It uses its own platform or job boards to source and shortlist CVs. “Last year, we hired 38,000 people and hardly 3% were hired through placement consultants. Our dependency on them has dramatically come down in the past seven years, from a 55% to 3%,” informs Anand.
Machine learning has helped with the internal recruitment for Soin at Tech Mahindra. As with many organisations, Tech Mahindra has a ‘talent pool’. Whenever, there was a vacancy requiring top talent, the company often looked outside for recruits. There wasn’t a scientific system to fill the demand internally. To bridge this gap, Tech Mahindra recently launched Talex, an AI-driven talent marketplace for its employees.
With this tool, they have mapped the skills of their existing talent pool. They mapped each employee’s profile, including their country, language skills, education and gender. “Once a job description comes in, the machine matches the profile and skills. If an employee’s profile has a 70% plus match with a vacancy, that job doesn’t go out (to a person outside the company),” says Soin. There are 121,842 employees in Tech Mahindra worldwide, so this system helps identify in-house talent, which would otherwise be ignored. That said, Talex was launched just three months ago, so its efficacy still needs to be tested.
If an employee’s profile isn’t a complete match for the vacant profile, but still aspires to apply, he can do so. The company has an interesting method for it, too. “We have given a dummy cryptocurrency to our members. They can bid for the job, irrespective of the match-rate. The highest bidder gets a wild card entry,” says Soin. It all starts with 100 points, but if an employee has done a great job in a corporate social responsibility activity or has earned a certification, he generates points for himself. Those points can then be utilised for bidding.
Tech Mahindra invests heavily in upskilling its talent for the future. As a result, it wants to use technology to keep them in the company for the longest time.
HCL Technologies also uses an AI-based system in its shortlisting and appraisal process; it gives them time-savings of a good 30-40%. “Over a period of time, we have invested in the algorithm, which helps us screen and shortlist right candidate based on scores. We score people on various parameters. Especially when people move from one process to other internally, this scoring algorithm helps us. If you overrule algorithm to take a decision because you have a different judgment, algorithm updates itself. It is a constantly learning system,” says Anand.
Advanced HR tech, which tracks employee behaviour and predicts it, is still new to most organisations in India, although a few sectors such as IT have begun adopting the practice (see: Gaining popularity). Paul Dupuis, CEO, Randstad India, says, “Because IT firms have a long employee trail and they have been collecting data for a longer period of time (as compared to other industries), they are in a position to leverage analytics and ML-based tech in the HR space.”
Data is generated throughout the employee’s stint at a firm. For instance, employee leave patterns, attendance patterns, reporting early, leaving early patterns and response to internal surveys, among several other parameters.
Peddi explains how predictive analytics work with an example. “I want to see, in the next three months, how many people will leave. I can look for common factors that can be identified in the past three to six months. For example, looking at metrics such as the amount of unplanned leave; less time spent on work and poor response. If there is a page where you submit your resignation, if someone has been visiting the page two to three times, most likely he might put in his papers. We track such visits also.”
HCL Technologies has been working on ‘predictive attrition’ for about two and a half years now. They have developed their model based on a variety of employee data. “We use a lot of information such as demographics, rating history, drop in ratings, the kind of grievances that have been raised, the kind of HR related queries, career related questions, leave taken, sudden leave applications and so on. We put as much data into the algorithm,” says Anand.
He says, “There are other layers in this data too. If the company’s attrition is at 15%, it may vary between genders, age groups, locations, and skill levels. So, these variations in data are also run through our models.”
Anand says, “At least two years’ worth of data is required to arrive at accurate conclusions. We started with a prediction accuracy of 40%, and we have taken it 70-75%. Not that HCL Technologies’ attrition has dropped in these years, but we have been able to keep it lower than the industry level and have been able to predict more accurately at an individual level.”
Since they collect data on external hires at the stage of the interview, they use that data to reach intelligent conclusions too. Typically, this data comprises gender, age profile, distance from work, earlier jobs, skills and so on. Anand says, “Some of this data has predictive value. We have arrived at a conclusion that, if you don’t make an offer for certain profiles soon after the interview, the acceptance rate starts dropping.”
Different companies may have different reasons for adopting HR tech, but there are some common factors: cost, efficiency, and consistency in decision-making.
As per a study by TeamLease Services, the advantages of incorporating HR technologies into business models are substantial (see: Changing times). Using HR tech can result in time and cost savings of upto 25-30%. Meanwhile, automation can save time and costs of performance management and training by upto 45%; talent acquisition, retention, and engagement by upto 55%; and employee productivity by upto a whopping 65%.
Monjin is one such start-up which has 82 clients globally and has crossed 2.5 million transactions over its lifetime. The start-up facilitates video interviews of candidates, taken by its panel of subject matter specialists. The interviews are then supplied to companies, instead of plain CVs.
“Overall costs come down significantly. $1200 was the cost-per-hire for some of our clients; they had to call potential recruits to a venue earlier. With remote interviews, the cost has come down by 62% in the first year,” says Abhijit Kashyape, co-founder and CEO, Monjin.
The average turnaround time has also come down with video interviews. Turnaround time is down by 50%,” claims Kashyape.
HR heads feel it is not really the cost cutting which is on their minds when they adopt HR tech. It’s mostly the efficiency, transparency, and speed. Panasonic has automated its old HR systems over the past four years. “Attrition is down from 18% to 11%. Average hiring time was 45 days, which is down to 21 days,” says Mishra.
Clearly, the adoption of HR tech is moving in a direction where the conventional methods of doing things are going to become extinct very soon. “I have a single slide dashboard today, completely driven by AI,” says Soin. It allows him to drive an organisation of 121,000 people. That’s the magic of technology.