Supply chains are one of the most underappreciated marvels of modern society. Finding the best possible solution to a supply chain problem is extraordinarily difficult, these are problems that can take computers days or even years to solve without the right algorithms.
Mustafa Vora, our next pathbreaker, Senior Operations Research Scientist at Lyric, a SaaS company, works at the intersection of several things — developing algorithms, building mathematical models, and working on the optimization engine that powers the platform.
Mustafa talks to Shyam Krishnamurthy from The Interview Portal about his direct PhD at IIT Bombay where he worked on developing cutting planes, a technique used in optimization to help solvers eliminate bad solutions faster and hone in on the right answer more efficiently.
For students, chase deep understanding over surface knowledge. When you truly understand something — not because you memorized it, but because you wrestled with it, sat with it, and worked through it yourself, it becomes a part of you.
Mustafa, Your background?
I grew up in Ahmedabad, Gujarat. My early childhood was spent in a close-knit nuclear family with my parents and grandparents, and as I grew older, I transitioned into a vibrant joint family. That shift taught me a lot about adapting to different personalities and navigating complex dynamics, life skills I didn’t fully appreciate until much later.
Academically, I was always strong in school, particularly in mathematics. From a very young age, numbers felt natural to me in a way that words didn’t. I was genuinely passionate about solving math problems, and that passion spilled over into board games and puzzles — activities I still enjoy whenever I find the time. Looking back, I think that instinct to break down a problem, find patterns, and arrive at a logical solution was always at the core of how my mind worked. It was an early signal of the kind of work I would eventually find myself doing professionally.
Interestingly, while I excelled at math, communication and language were a real challenge for me growing up. My writing was weak, and English felt largely out of reach during my school years. I was far more comfortable in the world of numbers than in the world of words. In hindsight, I think this contrast actually pushed me further toward quantitative and analytical thinking — it became my comfort zone and eventually my career.
What did you do for graduation/post graduation??
I completed my B.Tech in Mechanical Engineering from Nirma University, Ahmedabad, where I was a day-scholar. My original inclination was actually towards pure sciences like studying Physics or Chemistry since everyone around me was studying for engineering, but my family steered me towards engineering, which was the more practical path in their eyes. As for choosing a branch, I honestly didn’t have a clear strategy. I picked Mechanical Engineering simply because it felt like the most ”hardcore” form of engineering at the time. I even consciously avoided Computer Science because I thought, quite naively, as I would later realize, that it wasn’t ”real” engineering. That was a misconception I would laugh about many years later.
After graduating, I worked for two years at a supply chain company. That experience turned out to be a pivotal chapter in my life, as it exposed me to the world of Operations Research, a field that sits at the intersection of mathematics, logic, and real-world problem solving. Something clicked for me there. It was the first time I felt like my love for mathematics had a meaningful, practical application. That realization led me to pursue a PhD in Operations Research at IIT Bombay.
What were some of the key influences that led you to such an offbeat, unconventional, and unique career in Operations Research?
My path to this career was anything but linear, and I think that’s what makes it interesting. It actually starts before engineering, with my Class 12 English teacher, Mrs. Mini Mullath. At a time when I was someone who struggled deeply with language and communication, she took the time to work with me individually and helped me learn how to articulate my thoughts clearly. That might sound like a small thing, but it turned out to be foundational. Years later, during my PhD, whether I was giving a presentation, defending an idea in front of my advisor, or writing my thesis, that ability to communicate my thinking was what made the difference. She was, in every sense, my first mentor, even if I was too young at the time to recognize it. Then came two years of working at a supply chain company after graduation. That period was where my real intellectual curiosity took shape. I started going deeper into Operations Research on my own, reading books, studying higher mathematics in my free time, and teaching myself programming. I would spend evenings solving math problems and then writing code to verify or explore them. It was entirely self-driven and I loved every bit of it.
That eventually led me to apply for a PhD in Operations Research at IIT Bombay, the only programme I applied to because it simply felt right. I appeared for the GATE exam, barely cracked it, but I prepared seriously for the interview. Having already spent two years working hands-on in OR, I could speak from real experience and that made all the difference.
My PhD advisor, Prof. Ashutosh Mahajan, turned out to be exactly the kind of mentor I needed. He was supportive, intellectually generous, and pushed me in the right direction. I came in with a Mechanical Engineering background and self-taught Python, and he handed me a codebase written in C++. It was intimidating at first, but with his guidance and a lot of hard work, I got into the details quickly and started contributing meaningfully. Those years genuinely made me fall in love with mathematics, programming, and problem solving in a way that has never left me.
How did you plan the steps to get into the career you wanted?
My first job after graduating was at Indus Momentus Business Solutions, a company that provided end-to-end sales and operations planning solutions to medium-sized supply chains in diverse industry sectors like pharmaceutical, electronics, manufacturing, etc.
I spent two years there, primarily working on Advanced Planning and Scheduling implementation for a pharma client. We implemented a tool that will automate production planning for the plant and also schedule (hour-by-hour sequence of operations) products on each line based on the demand. Along with that I was also involved in maintenance and support for another pharma client and some solution development for an electronics manufacturing company.
It was my first real encounter with how mathematics and logic could be applied to solve genuine business problems. Supply chains are messy and complex by nature, and seeing how optimization techniques could bring structure and efficiency to that complexity was genuinely exciting to me. This wasn’t textbook OR, it was real, with real constraints, real clients, and real consequences. But the more I worked on these problems, the more I wanted to understand what was happening under the hood. I wasn’t satisfied just applying techniques, I wanted to know why they worked, where they broke down, and how they could be made better.
Tell us about your career path. Why did you opt for a PhD?
I also knew that a PhD is a very life altering decision especially since people only think of masters first. A PhD is not the most obvious next step, and for most people a masters feels like the safer, more natural progression. But my path was a little different.
By the time I decided to pursue a PhD, I had already spent two years working in the industry. That experience gave me the clarity that I wanted to go deeper. I had seen Operations Research applied in the real world, I had taught myself the theory in my spare time out of genuine curiosity. A masters felt like an extra step that wouldn’t give me what I was actually looking for. A PhD, on the other hand, meant dedicated years of focused research, which was exactly what I wanted.
In terms of eligibility for a direct PhD at IIT Bombay with a B.Tech degree, you need to qualify GATE, clear a written test conducted by the department, and then appear for an interview. The interview is where your genuine interest and preparedness really matter. As I mentioned, my industry experience gave me a strong foundation to speak from during mine.
It is not the conventional path, but if you have the clarity and the drive, it is absolutely worth considering.
That curiosity is what eventually pulled me towards a PhD.
When I joined the PhD programme at IIT Bombay, my initial inclination was still towards applied OR, building models, solving industry specific problems. But as I got deeper into the theoretical and computational aspects of optimization, something shifted again. I found myself drawn to the harder, more fundamental questions — not just how to solve a problem, but how to build the machinery that solves problems.
That led me to work with Prof. Ashutosh Mahajan on the development of Minotaur, a solver for Mixed-Integer Quadratically Constrained Quadratic Optimization, part of the open-source COIN-OR project. To give you a sense of what that means: imagine planning a road trip where you want to minimize fuel cost and travel time simultaneously. The relationship between speed, fuel consumption and distance isn’t a straight line, it’s curved and complicated. On top of that, some decisions are binary: you either take a highway or you don’t, you either stop at a city or skip it. Now imagine millions of such interacting decisions, all tangled together. Finding the best possible solution to such a problem is extraordinarily difficult, these are problems that can take computers days or even years to solve without the right algorithms. My work was to develop those algorithms — specifically for presolve, cutting planes, and branching techniques that intelligently reduce the problem size, eliminate bad solutions early, and guide the search in the right direction. Building these algorithms requires deep mathematical insight, and even small improvements can have dramatic effects on how quickly a solver finds the answer.
Alongside this, I was also involved in something much more tangible, developing the semester timetable for IIT Bombay. It sounds straight forward until you realise the scale of the problem. Every professor has preferences, courses have priority over certain time slots, rooms have capacity constraints, and no two important courses can overlap for the same set of students. I worked on both the optimization model behind it and the web tool used to implement it. It was a great reminder that OR isn’t just theoretical, it quietly powers systems that thousands of people interact with every day without realising it.
One experience that also stayed with me was teaching Operations Research to Indian Naval officers. It was a completely different kind of intellectual engagement — these were seasoned professionals who wanted to understand both the theory and its practical implications. Explaining abstract optimization concepts to people who would potentially apply them in high-stakes real-world scenarios sharpened my own understanding in ways I hadn’t anticipated. It was a reminder that the best way to truly know something is to teach it.
How did you get your first break?
My first break came through a campus placement at Nirma University. Indus Momentus Business Solutions approached the university looking for suitable candidates, and when the opportunity was posted on the job portal I applied and went through three rounds of interviews before being offered the role.
What made this moment interesting was that at the same time I had also received an offer from a natural gas distribution company as a Graduate Engineer Trainee, a much more conventional path for a Mechanical Engineering graduate. On paper it was the safer, more expected choice. But something about the Indus Momentus role felt more intellectually challenging and I chose that instead. Looking back, that decision set the entire trajectory of my career in motion. Sometimes the less obvious path turns out to be exactly the right one.
What were some of the challenges you faced? How did you address them?
Every meaningful journey has its share of hard moments, and my PhD was no exception.
The first challenge was purely technical. When I started working on Minotaur, I was stepping into deep water. I had no prior knowledge of how complex a solver’s technical stack really is, I was unfamiliar with the underlying algorithms, and I had never written a line of C++ in my life. It was overwhelming at first. But I would constantly learn new skills and read as much as I could on the topics I wasn’t aware about.
The second challenge was more subtle but equally important. During my PhD I developed what I believed was a promising conjecture about a particular class of cuts I had worked on. I spent months trying to prove convergence, going down every path I could think of. Eventually, after many discussions with my advisor, we concluded that this might simply be an extremely hard problem to prove and decided not to pursue it further. That was a difficult moment. I had invested so much energy and excitement into it. But it taught me one of the most valuable lessons of my career: not every battle can or should be won. Knowing when to stop is just as important as knowing when to push harder. In research, and in life, that judgment is everything.
The third challenge was personal. In the later years of my PhD I got married and had a child. Navigating the responsibilities of a growing family while living on a PhD stipend and pushing through the final stages of research was genuinely hard. But with my wife’s support I found a way through it, and if anything it gave me an extra sense of purpose to finish strong.
Tell us about your current role. Where do you work now?
I currently work as a Senior Operations Research Scientist at Lyric, a SaaS company that provides a platform for modelling and solving a wide range of supply chain optimization problems. In simple terms, we help businesses make better decisions about their supply chains by giving them powerful tools to model their problems and find optimal solutions.
My work sits at the intersection of several things — I develop algorithms, build mathematical models, and work on the optimization engine that powers the platform. The problems we deal with are rooted in real supply chain challenges, things like planning, scheduling, and resource allocation, where the decisions are complex, the constraints are many, and the stakes for getting it right are high.
What are the skills required? How did you acquire them?
At a very high level, both Indus Momentus and Lyric operate in the supply chain space, so I understand why they might seem similar at first glance. But in practice they are quite different companies.
Indus Momentus was a service and project based company — we would go into a client’s environment, understand their specific needs, and implement planning tools for them. The work was hands on and client driven. Lyric on the other hand is a product company, which changes the nature of the work fundamentally. We are building a platform that needs to work across a wide variety of supply chain problems at large scale. That requires a level of innovation, research, and algorithmic rigour that goes well beyond implementation.
In terms of how my work has evolved, the PhD made an enormous difference. At Indus Momentus I was largely applying existing tools and techniques to solve client problems. At Lyric I am designing algorithms, developing new approaches, and contributing to the product itself. The problems are more complex and more research grade. There is significantly more autonomy, I own solutions end to end rather than executing a defined scope of work.
And yes, the role is considerably more strategic. At Indus Momentus the question was often “how do we implement this for the client?” At Lyric the questions are deeper, what is the right algorithmic approach for this class of problems, how do we make it scale, how do we make it robust? That shift from execution to thinking about the bigger picture is something the PhD prepared me well for. It taught me not just to solve problems but to think carefully about how and why a solution works.
In terms of skills, this role demands a fairly unique combination. You need strong mathematical modeling ability to translate a messy real world problem into something a solver can work with. You need algorithm design skills to make solutions efficient at scale. You need proficiency in optimization tools, strong programming skills, and genuine domain knowledge of supply chains because without understanding the business context, even a technically correct model can be practically useless.
What I love most about this job is the mix of all these things. A typical day involves meetings with the team, working through problem formulations, and a lot of heads down coding, designing algorithms, and problem solving. For someone who has loved problem solving since childhood, it is hard to imagine a better fit.
How does your work benefit society?
Supply chains are one of the most underappreciated foundations of modern society. The food on your plate, the medicine you take when you are sick, the phone in your hand — none of it reaches you without a supply chain working correctly behind the scenes. They are invisible when they work and catastrophic when they don’t. The COVID-19 pandemic was perhaps the starkest reminder of this — when global supply chains were disrupted, the consequences were felt by everyone, everywhere.
But a supply chain is only as good as the decisions being made within it. When should you produce? How much? Where should inventory be held? How do you route deliveries efficiently? These are not simple questions, they involve thousands of interacting variables, tight constraints, and real consequences for getting it wrong. This is where optimization and algorithms come in. They are the silent engine that allows supply chains to plan better, waste less, and respond faster.
Better optimization means less overproduction and less spoilage, which matters enormously in food and pharmaceutical supply chains where waste has direct human consequences. It means medicines reaching patients on time, essential goods being available when needed, and resources being used efficiently in a world where they are increasingly scarce. It also means fewer unnecessary truck journeys, better utilized warehouses, and a smaller environmental footprint.
The work of developing better algorithms for these problems is therefore not just an academic exercise, it is about making the infrastructure that quietly powers everyday life a little more reliable, a little more efficient, and a little more resilient.
Tell us an example of a specific memorable work you did that is very close to you!
One piece of work that will always stay with me came from my PhD research. I was working on developing cutting planes, a technique used in optimization to help solvers eliminate bad solutions faster and hone in on the right answer more efficiently.
What I find most memorable is how it came about. I wasn’t following a grand plan, I was simply sitting with a small problem, solving it by hand, when something unexpectedly worked. I brought it to my advisor and over the following weeks we realized that what I had stumbled upon was actually a general class of cuts that were computationally cheap to obtain and mathematically guaranteed to always work.
That accidental discovery taught me that some of the best ideas emerge quietly when you are deeply engaged with a problem and curious enough to keep prodding it. The simplicity and elegance of the result made it all the more satisfying.
Your advice to students based on your experience?
The most important thing I have learned, and the advice I find myself coming back to, is this: chase deep understanding over surface knowledge. When you truly understand something — not because you memorized it, but because you wrestled with it, sat with it, and worked through it yourself it becomes a part of you. It stops being something you studied and becomes something you know. And the clearest sign that you have reached that depth is what I call the ”aha” moment. That sudden flash where something clicks, where you feel like you discovered it yourself. It doesn’t matter if thousands of people have discovered it before you, that feeling is real and it is yours.
That is the kind of understanding worth chasing. And here is what I have noticed: the subjects and problems that naturally give you those moments, the ones where curiosity pulls you forward without any external push, those will quietly become your career if you let them.
Future Plans?
I think the need for optimization as a tool is growing rapidly. More and more businesses are moving to optimize their decision making. I would continue to work on developing algorithms that power this decision making. With the advent of generative AI our field is also seeing some transformation, I am curious to see where it leads and what will my job get transformed into.