IIT Bombay at Inter IIT Tech meet 12.0

Introduction:

When the nation’s brightest minds unite, the confines of established knowledge fade, paving the way for unexplored realms and pushing ahead the boundaries of discovery. As they assembled at the Inter-IIT Tech Meet 12.0, they did just that, standing upon the Shores of Discovery. This year’s tech meet occurred at IIT Madras from 19th to 22nd December. 

The Tech Meet’s goal is to involve students in solving real-world challenges presented by companies and organisations. These challenges/problem statements (PS) fall into three main categories: High-Prep, Mid-Prep, and Low-Prep events covering diverse areas like machine learning, quantitative trading, consulting, robotics, and more. The High-Prep problems are announced as early as two months before the Tech Meet, while the Low-Prep challenges are released just 5-10 days before the showdown.

This year, alongside High Preps, Mid Preps, and Low Preps, there was a new addition: the on-the-spot No Prep challenge called Maths Bowl. This offline math competition followed the format of esteemed Olympiads like INMO. It promised participants an intellectually engaging and impromptu contest, bringing a fresh dimension to this year’s event.

The Upwards Trajectory And IITB’s Past Performances : 

IIT Bombay’s best performance came in the Inter-IIT Tech Meet 9.0, hosted virtually by IIT Guwahati, where we stood 2nd, bagging three gold, two silver, and four bronze medals. 

We have improved our performance over the past three years, progressing from a 5th-place finish in the 10th edition to securing the 4th position in the last meet to the podium finish this year. This year, we finished in third overall position, with three gold medals in the Aptos Blockchain Challenge, Zelta Labs Crypto Trading PS, and the Maths Bowl Challenge, respectively. One silver medal in the Trumio – University Project Ecosystem Challenge and one bronze medal in the Jaguar Land Rover PS. 

Gearing up for the 12th Edition:

As we buckled up for the upcoming meet, Insight talked to a few team members and Contingent Leader Balasubramanian about the learnings from past performances, preparation levels, strategies, and expectations from this edition. 

Here’s our interview with the Contingent Leader Balasubramanian:

First things first, let’s start with the selection of the team members for the contingent.

“We have a process in a place developed over the past few years. The ITC managers, GSTAs, and the Project and Tech Team Nominees maintain a database that keeps updating. It has information on who’s working on what domains and what competitions have been organised by the different clubs. Besides that, tech teams are there, working on a niche problem statement or for a competition, and they will be very proficient in that. Depending on the subsystem, we have a lot of people to choose from, with a stamp of working for a large number of hours on a single problem statement. Finally, the Contingent Leader itself is selected based on their technical and management skills, involvement in previous Inter-IIT Tech Meets and their network. I personally have a good network in the technical teams, department mates and also on the ITC side. 

We do send out a form to select a team lead, but it’s more dependent on personal reviews and peer reviews because a simple interview won’t suffice to judge how much they can contribute as a team lead working on a problem statement. So, the team leads are preferred to have people who can attest to their past work.

We also float a form to get interested people to participate in the meet. So, we do get many responses for many of the problem statements, but for niche problem statements, we get very few responses. Once the team lead is selected, we assist them in recruiting and assembling the rest of the team. 

The selection of the team leads also depends on the type of problem. Usually, once a problem statement is released, we try to finalise the team leads within a day or two. For example, for a low-prep problem statement, we finalise the team lead within 24 hours of the PS release, as they are mostly niche domain problem statements, so very few people are working on it. Meanwhile, for a high-prep problem statement, we might take two days, as 2-3 team leads might be required to work on it. 

Time does go into the selection of the leads and the members. They might also have other time commitments, and if someone leaves in the middle of it, then it is very difficult to find a substitute. Selecting the team leads is a very crucial part; they have two things to do: 

decisions on the path of the technical solutions which is being done (overall bird’s eye view)

and getting the work done by the other members of the team; each one has their strengths , and each one has their quirks. It involves finding a balance that works for everyone, and you need to know how to deal with them, and it is quite tricky when it comes to tech. 

The Contention System 

In the tech meet, there is a system to raise queries or attack other IITs’ presentations to have penalties imposed on them for not being able to defend the queries. 

This is different compared to the cult or the sports meet. In Cult, the judge’s decision will be the final decision, and comparing two art forms or performances depends a lot on the subjectivity of the judge. Over here, it’s different. Since it’s a technical event, technical and logical answers must also be defended, along with the queries or contentions raised by other IITs. 

Once all the contentions against an IIT are read out, a representative from that IIT gets a chance to answer these queries. If they are not able to answer them, then according to the rules as decided by the organising team and the other IITs and based on the severity of the contention, a penalty is imposed on that IIT. 

“Last year, we lost quite a few points because we could not defend a couple contentions against us. So we had kept that in our minds this time around.” 

About any new strategies being put in place to better prepare the contingent for defending the contentions raised, He said, “The technical queries raised depend on the particular problem statement and the approach we took to solve it. But general things that are to be kept in mind while participating; for example, in the tech meet, each IIT is assigned a random number, and that is the only thing that should show up in all the submissions; nowhere can we have the name of our IIT, this is done to avoid any decision bias. So, a lot of contentions arise from simple things like the mention of IITB in the submission. Therefore, multiple checks were implemented to avoid such naive mistakes.”

He also added: “There were problems in the past because of lack of coordination in the ITC and the Inter-IIT Tech Contingent; additionally, the roles and responsibilities were not clear enough, leading to several decisions being overridden by many people. This year, to avoid that, we had the distinction of roles between the CLs, CM (Contingent Manager) and the different PoCs quite clearly stated in the beginning. Also, efforts were taken by both sides to improve the coordination. The financial and logistical support that we have received from ITC and the GSTA has also been very good.”

Q. How have the problem statements changed over the past 3-4 years?

→ 

“Over the past three years, the difficulty level for problem statements from particular domains has remained quite the same. But you can’t compare problem statements from different domains, so gauging the overall difficulty trends is hard to do. 

Speaking about the type of problem statements. So the Inter IIT Tech Meet 9.0 (held in 2021)  was completely online, and that’s when, understandably, we saw a huge shift towards more PSs from Software Dev, simulation, and AI/ML, which then became a trend that has continued so far. Not surprisingly, we had only one hardware problem statement last year, the Drona PS, which continued this year; we had only one mechanical PS, and the rest were software-related problem statements.

So this trend is really demotivating for many Mech enthusiasts. So, many students are interested in mechanical design and robotics. Still, sadly, for the past 2-3 years, Inter IIT has failed to provide these students with the right set of problem statements. So, that is a worrying thing to observe.

But apart from that, the software problem statements have improved and become more interesting. PS’s intersected across domains such as cybersecurity and DL where given which are the kind of problems we live to solve. They are also relevant to the current trends in the industry. Last year, they had an NLP-based problem statement; this year, they had an experimental problem statement from the domain of LLMs, which was very interesting and good to see.

At the same time, problem statements like WorldQuant, in my opinion, it is a very narrow problem statement. The problem statement was to create more alphas [mathematical models that seek to predict the future price movements of various financial instruments] and was put as a high prep problem statement. Now, here’s the thing: the company already has a WorldQuant BRAIN Program in place, where they seek to generate alphas and other algorithms. So it was irritating to see the same problem statement; they have the program and similar competitions, and they still gave the same PS. So, that is not something that you want in the Inter-IIT Tech Meet. There is no real value addition happening to the tech community through problem statements like these; only the company benefits by getting a bunch of more IITian consultants great at creating alphas and trading algorithms.” 

Q. What lessons have you learned transitioning from participant to leading the contingent across three tech meet editions?

→ 

“In my second year, back in Feb 2022, in the 10th edition of the meet, I participated in the DRDO UAV UGV Guidance and Navigation challenge through my tech team (UMIC), as it was a problem statement relevant to my tech team’s work. They just gave it to us, and we solved it. 

But that year, I couldn’t see the picture on an overall scale, how much tech is pervasive in the institute, or how much liking there was for it. 

And we were quite disappointed when the standings came, we finished fifth that year. After all these efforts, we worked properly day and night for over a month on that problem statement, and we got a bronze medal. So that was a big learning for me. 

Next year, I had a different objective in mind. Because I wanted to target CL this year, last year, I wanted to see the management side of things and the difficulties that come with it. So, I became a core team member. They are not directly involved in the tech. Still, they are in charge of getting updates from the participants of the problem statements the team leads and directly updating the CLs and the CMs about the situation over there. They arrange all the logistical help for the team. So they are like the Seargents of the team. It was quite a rewarding experience. This gave me a lot of exposure and also a connection to my previous CLs, Devank and Devansh. We had a lot of discussions on managing people, taking care of submissions, and dealing with technical and non-technical situations where you don’t really have any idea about it.

Being a team member, you focus on the PS and what part of the problem statement we are asked to work on. Dealing with people is a different game; it is a game of navigating a delicate balance between assertiveness and cooperation, often described as the give-and-take policy. This balance is crucial in fostering effective communication and collaboration within a team.. It is all about conversations. I feel like this position, or any other leadership position, helps you overcome your own insecurities. When you are in a team, you realise one fact loud and clear “it is not about you”; it is always about the team and the collective results. Yes, you will be put into uncomfortable situations and conversations, but you will still do it as it is necessary for the team. So, it gives you the greatest courage.

Many things must be sacrificed, and the right kind of choices are to be made to fuel the team. It is an interesting job, understanding the peoples’ needs and managing them, adding value and also getting value out of the team.”

Q. What areas of improvement can we as an Institute work on to improve our results in the Inter-IIT Tech Meet and, in general, to increase the tech drive in the student community? 

→ 

“This competition is about passion and glory. It shows us the ground reality of tech in different IITs, and this is where I have a few complaints and concerns, which are also shared by most of the previous CLs and a few GSTAs. But, we are trying our best to improve the situation and bring awareness to the tech community in the institute. 

Kharagpur, Roorkee and Kanpur have the strongest tech, and we can see that in the standings of the last couple of meets itself. Last year, we finished fourth only after these teams, KGP, Roorkee, and Kanpur, and the point difference between us and them was huge. So, a lot of effort is being put into it, and they take great pride in participating in these meets. 

In our institute, I would say there’s a culture of taking great pride in participating in Inter-IIT cult or Inter-IIT sports. Still, when it comes to Inter-IIT Tech, many people are hesitant to participate, even though they are really really good at tech. But that culture is slowly changing, in my opinion, at least from what I have seen from participating in the tech meet for the past two years.

A lot of effort needs to be put in. We are in IIT Bombay, and we do have the best brains, but that itself is a problem; the “this is enough” kind of attitude sets in, and our culture of submitting assignments late at the last second also comes directly along with it. So, the output for last week will be brilliant compared to the output we will see for the last two months or so for a high-prep problem statement. 

It is good at times, but when it comes to competition, where other IITs are putting in these kinds of efforts on a daily basis, that’s not enough. At the end of the day, this is a competition. 

The problem is difficult to dissect, but I will try my best. I am one of the OCs of UMIC, and in my technical team, there’s a structure put in place; most of the technical teams and clubs run in the insti and are very well run. Some convenors are ready to put down their life. But the same kind of dedication when it comes to doing Inter-IIT tech is not found. similarly, one major difference between Inter-IIT cult and Inter-IIT sports is that people in the clubs are the first preference for the Inter-IIT. But in Inter-IIT tech, you can’t do that because you don’t know what kind of problem statements would come at that particular time. 

In sports, you have a team or a group practising a sport for a year, and they just need to select the best bunch out of the camp and send them to the competition. In cult as well, you have a few people who have been doing it over the years. For us, the entire institute is a pool of candidates to select from, and we need to select them every year again and again and again in a span of less than 48 hours!

Most clubs want to cater to a large audience, and their goal is to ensure that their workshops and events serve as a stepping stone for people who wish to explore tech. To pursue and push further [into their respective domains] is not what many clubs think of, I would say, and bringing that in will be a very big advantage to us. A lot of efforts need to be put into correcting this, and we need people who have that level of technical expertise and are also willing to impart that knowledge to others and spend time in doing that, and that is a very rare thing as it’s a lot of effort, other IITs do that.

That kind of forms a senti, and it is passed on from one batch to another. IIT Roorkee has a very strong coding community compared to ours in my opinion. Over here, we do organise few coding competitions here and there, but to run the club purely for competitive programming or coding and to maintain that drive to dive further into the subject and get better at it is a hard thing to do. and we do fail over there.” 

Contingent Leader Rohan Kalbag also gave his inputs on areas of improvements and concerns: 

“From the example of IITKGP which has a knack of coming first for the past few years. As a CL attending a lot of their presentations and observing their teams a few practices we (IITB) can follow.

Equal importance given to all problem statements, be it a high prep or a low prep their presentations were of equal quality. Their Panchayati Raj (last low prep) ppt was at par with the JLR ppt (second high prep), from the quality of the presentations it seemed as though the slides were not created a few days prior to a day before. Many days or months had been put into these slides, most followed a light professional theme, with lots of visualizations such as images, plots, simulation results and business proposal (how the company can inculcate this in their existing products or codebase) these carry very significant value and emphasis must given to the novelty or how our solution differs from other ideas, filler slides one the technical gruesomes, notation and definitions should be avoided since there will be other IITs who will present this and judges get bored.

Need for oratory skills, our guys do really well at showing our technical skills, but do a mediocre job at conveying its contents to the judges, it seemed as though each PS of KGP had 1-2 people specially for the presentation who would fluently convey the information to judges in the specified time limit at a good pace. Taking the example of our Trumio team, having people with oratory skills gave us the edge, our presentation was something like out of an new Apple iPhone release, the judges were rocked.

During the QnA, another point is to wait to structure the content and decide amongst all who is the most comfortable in answering it. Beating around the bush gives a bad impression to the judges. One line bang to the point answers like our JLR guys should be what we should aim for. Having a lot of mock QnAs before the actual ppt helps a lot here.

Discipline and dressing ethics, be it a high prep or a low prep all the KGP guys were in tuxedos and well groomed giving a professional touch. It’s the small points which make the difference. Even in the online ppts, the KGP guys switched their cameras on with suits and presented their work.

Just acing at the presentation is insufficient since they are often restricted to 30% of the weight. From the technical side,  providing extra details and being as detailed as possible always helps. In JLR, the KGP team has done not only the microarch block design, they had done very elaborate simulations on gem5, spice, also detailed thermal simulations on ansys, these weren’t specifically requested for in the PS but would give a sense of completeness. Similarly in the Certin PS, there was no mention of a prototype to be created with implementations (website with implementation of proposed security strategies) our guys did a splendid job on describing the strategies and documenting it well, but the winning IITs implemented the strategies on a demo website and showed a working demo, which definitely gave them an edge. So doing things not specified in the PS, going into details makes a positive effect. Detailed peer review before submissions with experts in the field can help us achieve this better”.

Q.  How do our chances look this year, and what are your expectations from the contingent as a whole and your personal expectations?

→ 

 “As a whole, the contingent has done excellent work. There were some difficulties in selecting the team; we couldn’t get the most suitable members to work on some niche problem statements because of fests happening in the insti at the same time. Usually, fourthies are the best people to have on the team, especially those who have their placements sorted. They have nothing significant to do and are quite talented because of their varied experience. But the problem is it’s the last Mood Indigo or the last fests for them, and they wanna chill and enjoy with their friends, and this [Tech Meet] directly comes during that time when we have MI and Sports Meet happening at our campus. 

But irrespective of that, the people who did apply were extremely talented and dedicated. Even if they had a little less experience in some of the domains, their dedication and the amount of extra effort they were willing to put to drive the team to at least have a chance at winning the gold is quite visible, and you can see it in their work.”

Talking about a mid-prep problem statement from the domain of Quantum Computing concerning the scheduling and re-allocation of Airplane PNRs, He said, “The efforts put in by the team were incredible. Many of us, including me, my co-CL Rohan, CM Aakarsh and GSTA Prathamesh, couldn’t sleep because of the submission. Our team did a really great job at this problem statement today (day 1). It will continue tomorrow as well, and till now, we have a good advantage over the others, be it the technical aspect or the business aspect, all of it was taken care of by the team perfectly.” 

“Amazingly, you wouldn’t believe it, but the team comprised of a couple of fourthies and PhDs, but the rest were thirdies and sophies. I mean, these guys were doing quantum computing, which is just amazing. I thought these guys were fourthies, who were done with their placements and are now doing quantum computing, but this morning before the presentation, they were calling me ‘Bhaiyya’, and then I got to know that these guys are sophies and thirdies. They also haven’t slept till now…such extreme dedication, and I believe if this persists, we have a good chance of winning.” 

“So yeah, the first day went pretty well. There have been slight changes in the rule this year, which has made a level playing field and has added more excitement to the competition. So, we do have a good chance this year. Obviously, as a CL, my job is to be hopeful, push the team to its fullest, and strive for the best results. But you can’t say anything about what position we will be in because anything can go wrong at the end moment.” 

Comments from other team members:

Kushal Agarwal, a team member working on Solinas Product design problem statement, had the following to say when asked about his problem statement; “It is a mechanical product design PS, given by the startup that aims to clean septic tanks and manholes to remove human scavenging, for that they wanted a telescopic shaft, which is basically a shaft that can expand and retract, and then a propeller can be attached to the shaft to clean the tank”.

He has also participated in the previous tech meet and had this to say about his learnings, “Being a part of the tech meet for the past two years, I can definitely see myself growing…Last time, I was heavily inspired by the amount of tech that goes on within Inter-IIT Tech Meet. Now it’s even more, where I am actually going offline and representing my campus..so it is a wholesome experience wherein you learn not just what tech is but how to actually deliver the tech that you’ve built over the past couple of months ”.

He further added, “You can see the spirits are high; all of them have been working day and night, and I can see each member dreaming of gold as we reach there”.

14