With over 1800 students, INSAID continuously delivers quality Data Science education to budding Data Scientists all across India.
Today, we are in conversation with INSAID’s Chief Operating Officer, Nikhil Bhogaraju, to see how INSAID achieves the best classroom environment, competes with global standards of learning and promises aspiring Data Leaders to flourish in their domain.
Malvika: Can you share the vision behind INSAID’s structured curriculum and programs?
Nikhil: The biggest goal when we were building these programs, whether it is Certificate in Data Science Foundation, or Global Certificate in Data Science or Global Certificate in Data Science and AI, was that we cannot have programs which are simply meant for people to learn, they need to have a clear objective and be curated as outcome-based programs.
What I mean is that different people have different levels of motivation or available bandwidth of time because 99% of our students are working professionals. Thanks to these variables, you just can’t put everyone in one basket. You can’t say that to learn data science from A to Z, you’ll only need one year of time frame!
What happens with this is:
First, people will lose motivation over a period of time.
Second, the program completion rates will fall big time.
Third, there will be no success stories around
Lastly, people will not be able to get what they want.
So we have tailored the programs in such a way that one is a short duration program of 3 months, which will get people introduced to concepts of, say exploratory data analytics or focused on data analytics and basics of machine learning. This is good from a senior’s perspective who just wants to apply analytics so that the outcome of the application of analytics is very clear within the three months time frame.
Now people who are a little more focused and they would want to transition into this industry altogether; those who would want to work day in and day out with data, they need a definite 6-8 months of time frame to learn things and actually do projects around it. In this six months bracket, we would also not want to squeeze everything obviously. At the same time, we don’t want to stretch things also, equally.
So between the 3 months program to 6 months program, our approach is first centered around the programming language, which is Python. The students are already learning the language and it is this one which the entire curriculum is based on.
When they get towards the last term, if I were to specifically talk about the GCD program, they get the flexibility to choose electives and go deep into what they want to do.
For example, if they want to expand their skill set of tools, they can learn programming languages including Python or R or they get to choose visualization techniques.
The second route they can take is that since they’re done with supervised and unsupervised machine learning algorithms, they can now choose to go deep into machine learning with Advanced Machine Learning. So that’s the goal of a 6-month program.
Now when it comes to the 1-year program (GCDAI), that’s where we have placed major bets just like the entire industry has. We know that by probably around 2022-25, there is going to be a terrific AI boom that has already starting to catch on but year on year, it will streamline and increase.
Right now in 2019, there are people who are early movers who would want to transition and would want to make the most in the next 5 years. They can’t start learning from scratch when the wave is at its highest peak.
For such aspiring Data Leaders, the AI part comes into the picture. Essentially this is 6 months of data science and machine learning and 6 months of AI where you’re focusing on exploring deep into computer vision and NLP.
So that’s how these 3 programs are focused on what kind of outcomes from these three programs one should expect.
Malvika: INSAID students hail from different backgrounds. What are the various profiles that you get at INSAID?
Nikhil: Initially when we rolled out all of these programs and we started getting traction, one of my biggest hypotheses was that this is related to IT and the uptake for this program will majorly be from IT firms.
There are other firms obviously like this whole industry has started its roots from healthcare, but in terms of uptake of the program, I never expected that other streams will come into the picture.
To my surprise, you name a stream or function and we have a student from there. The distribution may be leaning towards IT firms but there are people who are from BFSI, Telecom, product, e-commerce, healthcare and consulting.
In terms of companies as well, you name the unicorns of startups, for example, Flipkart we have students from there. We have students from Consulting firms like BCG and McKinsey, employees from Product leaders like Microsoft, Oracle, Cisco, we have them all!
That is what excites me the most that we have students all the way from 1-2 years of work experience way up to 20+ years of experience. Students are looking forward to learn data science and actually implement data science differently based on their career stages.
They have different objectives; freshers would want to transition, they would want to get salary hikes and seniors who have 15+ years of experience and are already at directorial positions would want to set up Data Science departments, they would want to be those practitioners who are solving actual core business problems in their organization.
So that is the distribution of students. Over a period of time what I’ve also observed as a trend is that the number of people who are already ITians and are done with their MBAs have also started significantly increasing.
Initially, for example, out of the class size of 120, 2 people from an entire batch of 120 were from a different background but since the last 6-7 months across every batch almost there is a good mix of 10+ MBA students, 5+ IITians, 3-4 BITSians and the rest are from NITs who are taking up all these programs.
Malvika: How do you prepare students before their program at INSAID starts?
Nikhil: So before every batch starts, what we do is, the day somebody registers for a particular program, they get access to our starter kits.
Suchit Majumdar, INSAID’s Chief Data Scientist and Senior Faculty for machine learning, has himself prepared the starter kits on both Python and statistics.
The whole objective of these starter kits is to bring everybody to a common ground because there are non-programmers, programmers, managers or people from management backgrounds who are joining.
From day 0 to day 10 when things are totally confusing and people have just started and are not able to grab things, that time we don’t want people to have foundation level questions.
Ground-level queries like how to install Python and what is the difference between mean, median and mode and such basic concepts should be clear.
It’s not always that people don’t know, it is just that it is always good as a self-check for everybody or as a refresher.
Ideally, for someone to finish this starter-kit, it hardly takes any time. For example, an aggressive learner would finish it up within a week. Somebody who is just spending about 30 minutes to 40 minutes a day, would take about two weeks to finish these starter kits.
This is the most important thing that we always push to our students who are joining in the incoming program to finish their starter kits at the earliest.
Malvika: What checks and balances does the Academics team at INSAID have to resolve student queries beyond classroom hours?
Nikhil: Firstly, let me just put out a quick sense of how classroom queries are resolved and then I’ll talk about how queries beyond the classroom are resolved.
During classroom hours along with our faculty, we have adapted an offline model that along with faculty there are teaching assistants who are placed in the class and the whole goal of these teaching assistants is to resolve queries.
Now given the batch size of actual INSAID students who are attending classes, not everyone’s query is related to everyone. That is a known fact. Some might not understand, some will understand, someone’s query will be relevant for the entire class or some might be asking totally out of the box questions.
Whenever these kinds of questions are raised which are very specific to one particular candidate, these teaching assistants come into the picture and resolve their queries right then and there.
Whenever the question is beneficial for the entire cohort, the faculty takes up such questions live on air.
Another thing we do post classes, is that the faculty spends half an hour with the students where students come up on the air and get their doubts clarified after the classes. So this is how the classroom queries are taken care of.
The biggest challenge is that because this is a weekend program and it’s totally online, the students queries beyond classroom hours become a challenge. The way we have set up the Academics team is such that there are both technical queries and non-technical queries and these are dealt with by different sub-teams. Program-related technical queries are resolved by a dedicated research team who are the acting teaching assistants in classes. These researchers handle these queries.
So there is a mechanism in the website itself where the query can be posted and can be resolved at the fastest pace. If there are non-technical queries, which are program-related or field related or related to the program duration, there is an Operations team which is our non-technical team and will take care of that part.
The biggest benefit I would say is that we have come up with something called collaborative learning. Again, it’s not something that is not existing or entirely new but we have just brought our community into one place together.
In our community, we have streamlined things based on topics which is irrespective of batches, irrespective of the program a student has enrolled in.
For example, there is a topic like data analytics. Any questions you have on data analytics you can just post them over there and close to 2000 students (current strength of students) get to see your question and they start engaging with you and start resolving your queries.
On top of that, the research team is obviously active on the community. The way this entire setup benefits students is that firstly it is the fastest way to resolve queries and secondly because there are active participants on community if someone faces and entirely new difficulty, he can refer similar queries posted earlier by some other person.
So I think this is the biggest advantage that the fastest way the team resolves the queries is through the community.
This ends Part 1 of our conversation with Nikhil. Check out the rest of the conversation here!