This is Part Three and the final post in a series on the skilling structure at iMerit, where Chief People Development Officer Anindya Chattopadhyay shares his insights.
Over time, the focus within L&D shifts to up-skilling and cross-skilling. Upskilled employees gain deep domain expertise, move to more complex tasks, and become Quality Analysts. Cross-skilled employees gain breadth of domain expertise, can connect diverse data streams and ease organizational knowledge management. Leadership potential is nurtured through training in soft skills, communication, and floor/project management. In the past three years, 100+ inclusive employees have successfully progressed to team lead and managerial roles. Recently, through an innovative scheme of internal job postings, the L&D team itself hired 20 of our contributors and trained them to become process trainers and facilitators.
In the past three years, 100+ inclusive employees have successfully progressed to team lead and managerial roles.
While significant planning and ingenuity have gone into the design of our L&D programs, what gives us scale and agility is technology. Training programs are digitized and delivered in blended form through a Learning Management System (LMS) that supports anytime/anywhere learning, made engaging through gamification and personalization. Intelligent tools for assessment provide targeted feedback, helping practitioners improve their skills. For example, a home-grown CV assessment tool trained on ideal image annotations provides feedback on accuracy and eye for detail. It is regularly used to evaluate a user’s affinity for CV tasks. The employee develops into an agile and continuous learner.
Q: How do you motivate the team through the training process and thereafter?
Answer:
The company’s growing in leaps and bounds. So initially it was just me who tried to be on the floor and be friendly, and understand what their problems are. I even have so many small Whatsapp groups with employees from all of the centers – more than a hundred groups at this moment. So I can talk to them about their lives as well as the challenges in their work. We often need to take instant skilling decisions and recommend particular methods to them. I also follow up to see if this was helpful. This self-learning approach goes much deeper into the brain, where learning becomes a habit. This is very important at iMerit, because work is constantly changing and evolving. We started computer vision projects with a basic bounding box kind of level. Currently we are working on processes like LIDAR. It’s a 3D process where all the images are very dark. They need to really understand the spatial volume of that particular object from very different positioning of cameras. It’s a different kind of observation that they need to build up. Their eyes are now built at that level of observation. That is evolution, which requires the kind of growth in their knowledge.
Sometimes we have clients who give us time for a structured training opportunity. At this moment we have more than 350 people working on LIDAR processes for the automotive sector. Now that has given us the deliberate structured training opportunity.
We have trained people and assessed them. We have found that a percentage of the cohort is good to go to the floor, some percentage requires training. After they are on the floor, we check for quality. If there is a fall in the quality score, we help them again. But now we are expanding L&D, with some LMS analysis, and dashboards with data, which help us with decision-making and making ad-hoc situations replicable.
Now we have a small L&D team scattered among different centers and shifts and processes.
I train them to be a buddy on the floor, so that their presence can be felt by the team, where they can discuss project situations and there can be instant decisions. And then gradually that information is what we collect and put into the LMS. There is a bank with the knowledge, the courses, programs use cases and, the edge cases. Then there can be a kind of modular, course program where we can start certifying people for their skills at different levels. So we have X number of skilled people in LiDAR. So we can really go for that kind of scenario gradually.
I train them to be a buddy on the floor, so that their presence can be felt by the team, where they can discuss project situations and there can be instant decisions.
The best part of our social model is that motivation levels are very high and attrition levels are very low. These jobs are life changing for the youngsters. So the investment in L&D and the accumulation of knowledge and skill over time, is retained in the organisation and customers love working with stable teams. Even customers tell us that they like investing in training our teams because they see the fruit of their efforts over time.
Q: How do you instill principles like responsibility and accountability in a young workforce?
Answer:
Discipline is a requirement in an organization or in an individual. You can make people in an organization very disciplined If you have very strict procedures and you are a process-driven organization. Large organizations do work in that way and they are quite successful. One day iMerit may have a massive workforce and also follow the same thing. At the same time, large work organizations value educational qualifications a lot. Qualifications give them the idea that a person has gone through maybe 13 to 15 to 18 years of disciplined study. There is a discipline in their study previously and there probably will be a discipline expected from their work. Converting the discipline to study into discipline to work is not as difficult. So that works in many organizations.
However, iMerit has a philosophy of empowering youth from underserved communities where people do not have extended education, and they do not have any idea about the kind of job they would be doing. Then that basic skill, that discipline is not expected.
First, we need to make them disciplined individually, and then as part of a small team under the team lead, then small member of a large project, then maybe in the center and then the company. They are reasonably sincere when know that they are from this organization, from the center, from this project, and from this team. And this is where delivery team organization and hierarchy is pretty important. So we have a good number of people in the delivery team who continuously give this kind of information about how to work and about team effort.
Q: How do you customize your training method and your material for different individuals and different learning styles?
Answer:
I tried to create something that can be codified. Any coding requires a pattern. So identifying that pattern of what is needed for which segment of people can only be obtained when we start recording that in this segment we have done maybe two things, out of which one really succeeded and one failed. So then there is a pattern. There is a segregation of individual clusters and then a codification for that particular cluster. If a newcomer falls in those clusters, then we can try that particular approach for them, after certain kinds of psychometric tests to understand the person.
We have started identifying primary and secondary skills. Rather than seeing only one particular area or one particular way of training the person, we try to have a primary and a backup.
We have started identifying primary and secondary skills. Rather than seeing only one particular area or one particular way of training the person, we try to have a primary and a backup.
This Metiabruz team initially started working on projects like zoning and proofing of IEEE journals, where we needed to mark each and every paragraph and give them a category. Now that was a very text-heavy categorizing project, with a very primary level of tagging. At the time they did not do so well because of the English language understanding needed, though we tried to help a lot. We then found out that they’re good with images. When we started giving them image skills, they came out in an excellent manner. Today their comprehension has also progressed, so we are back to language skills and crossover of image and language tasks such as classifications and descriptions.
So there is research, tests, and trials, and we try to see if it is feasible. But to the next level within that image domain, whether the person is good for work in LiDAR, Skynet or pixel – that classification is something we are just starting to do. This is the second version of L&D that we are doing. After six months, perhaps we will be in a position to really say that this can be replicable and scalable through this particular structured manner. That structuring is actually getting done at this time.