Ever wonder what kind of statement of purpose impresses engineering admissions committees? I think about this every single day. I think about the structure. I think about the stories that Computer, Electrical, and Industrial engineers tell. I think about the style and the “Why This Program” section. But more than anything, I think about this one issue:
What kind of intellectual question does the applicant need to present?
“Intellectual question?” you might ask. “This isn’t philosophy. I want to work at a FAANG company. Don’t I just need to list my credentials, tech internships, and research experience? Don’t I just need the minimum scores?”
Having read thousands of STEM essays at both the master’s and PhD level, for applicants who’ve earned admission to every elite program a dozen times over, there is one thing of which I’m certain:
The best applicants are those who present the most interesting intellectual problems in their statement of purpose, then explain how they hope to explore them.
That’s why I was so fascinated to meet Haneesh on Reddit.
When he first contacted me, Haneesh’s credentials were amazing. BUT…he had the same problem that 99% of applicants from India have, and that many American students also face: his SOP was a chronological “autobiography.”
It started with a childhood story. (No! Don’t do this!)
It went on to his high school and undergraduate years. (Seriously, no high school stuff!)
It spent two paragraphs explaining his year-by-year professional experiences. And only in the last paragraph did it arrive at the present and finally mention the school he was applying to. (Sigh. These essays make me sad.)
I understand that in many countries these “autobiography” essays are ideal. But not in the United States. I also understand that many Western students get bogged down in their frame narrative. But in the end, these meandering essays are horribly ineffective and always result in subpar admissions at elite US universities.
Haneesh clearly hadn’t read Structure is Magic.
But there was a single sentence that piqued my curiosity. In this one sentence, Haneesh posed a big question. He actually seemed excited about it. It was deep and thoughtful. It talked about a specific opportunity at his target school, and it made me think: “Hmmm, now this sounds like a true scholar.”
I gave Haneesh some “tough love” advice. Structure it properly. Focus on that one unique idea. Then, like the gracious human being he is, Haneesh accepted the advice and went back to work.
9 months later, Haneesh messaged me again:
Hey Jordan, I got into a top-10 school! I structured my SOP just how you described. The key to a good SOP really is it’s structure. I can’t thank you enough! 🙂
Obviously, I was thrilled. But not nearly as thrilled as when I read Haneesh’s updated SOP, and saw the great pains he took to unpack that same, deep, intellectual question he’d barely touched in his first draft:
While trying to beat the stockfish chess engine, a question popped into my head: if we had a set with an infinitely large number of games, and assuming that both sides make perfect moves, could we prove that white always or doesn’t win? To put it more simply: can we make a winning decision from a set of an infinitely-large number of future scenarios? Certainly, I could extend this question to other games and real-life scenarios. Unfortunately, computing all probable states and representing the states as a graph is a time-consuming and expensive exercise. Stochastically optimizing these large graphs and churning out meaningful insights would require unorthodox approaches in optimization and Machine Learning. Yet, these difficulties represent exactly why I believe the MS in OR at Columbia University will provide the ideal environment to continue my education.
Working as a Machine Learning Engineer at Stark AI helped me understand how mathematically expensive it is to compute and optimize a Neural Network graph. Later, at Wayne Enterprises, I had the opportunity to manage and architect a Blockchain system, and was able to show how the consensus mechanism could be used to enable supply-chain transparency. Working for both of these startups helped me understand these challenges on an intricate level, while developing and maintaining large projects. After graduating from Eastern Univ. Engineering, I joined X-Consulting as an Application Development Associate, typically helping Japanese clients transform their business needs into technical specifications. Within days of joining X-Consulting, I took up voluntary leadership roles and delegated work to other Associates. Today, I am a part of a high performing agile team responsible for architecting and integrating SAP-ABAP analytics modules in large-scale SAP-ERP systems and data warehouses. Working for X-Consulting helped me understand the need for organizational hierarchy and the importance of data-driven decisions in an era that generates petabytes of information every second.
At Columbia University, I aim to improve my operational and technical skills, particularly in discrete optimization. Distinguished professors in Columbia’s Department of Industrial Engineering and Operations Research’s – Professor Charles Xavier and Professor Bruce Banner – have already been influential in my education. Their work on the Economics of Permissioned Blockchain Adoption allowed me to introspect the viable strategies required to adopt a permissioned blockchain network. Their thesis concludes by stating the strategies for adopting a blockchain so that the adoption is in equilibrium with the cost. In graduate school, I hope to study systems alternate to the blockchain that are inexpensive and can scale with ease. I believe without a doubt that by working with faculty such as Dr. Xavier and Dr. Banner, I can find an economically viable strategy that is as effective as a blockchain network, without compromising the consensus mechanism.
I plan to choose the Optimization concentration and include Computational Discrete Optimization in my coursework. After completing my master’s degree, I hope to work as an Analyst/SDE (Machine-Learning) for an organization like Google, Goldman Sachs, or Amazon. In the long term, I intend to become a Chief Technology Officer of a technology company. Thus, the department’s high job placement statistics (93%) are a major reason for my motivation to apply to Columbia University. I believe that my varied work experience and the coursework I have completed in the past few years have all prepared me to pursue an MS degree in OR. I am able to pick up abstract concepts and new mathematics quickly. For these reasons, I am confident that I would excel in the program, and that upon completion of my graduate degree, I will be able to solve optimization problems (np-hard/np-complete), architect robust Machine-Learning models, and make critical decisions driven by data. To be offered admission to Columbia University would not only be a tremendous honor, but also confirm my hard work in the past, and my commitment to work even harder in the future.
It’s funny, actually. This SOP doesn’t entirely conform to the template I advise in Structure is Magic or the Master’s SOP Starter Kit. Sections 2 and 3 – “Why This Program” and “Why I’m Qualified” – are somewhat inverted. Also, the final section doesn’t really “close” the frame narrative from the first paragraph. If I wanted to make this essay shine even more, I’d switch those two sections around. Then, in the final paragraph, I’d start with a short introspective thought about the stockfish chess engine and deriving insights from stochastically optimized data sets.
This essay does a wonderful job of narrowing down Haneesh’s intellectual interest. He wants to study unorthodox approaches in optimization and Machine Learning so he can glean insight from monstrous data sets. He knows how mathematically expensive these challenges can be. He knows how professors at his target program have approached somewhat similar tasks with blockchain networks. And most importantly, he knows exactly what he’ll gain from this program. He says so beautifully in the final paragraph:
“I will be able to solve optimization problems (np-hard/np-complete), architect robust Machine-Learning models, and make critical decisions driven by data.”
When we read this SOP, we know that Haneesh is singularly focused on one set of intellectual challenges. He has a pure academic focus, and by exploring this specific focus, in this specific program, with these specific professors, he’ll succeed in achieving his specific professional goals.
His credentials prove that he already knows how to succeed. Thus, when we read about Haneesh’s new intellectual interests, we know he’ll succeed in pursuing them as well.
Whether you’re applying to master’s or PhD programs, there are two great lessons you can learn from Haneesh:
Luckily, the SOP Starter Kits (for Master’s or PhDs) will teach you how to implement both of these lessons quickly and easily. Before you start writing your first draft, before you start pondering your autobiography, work through the exercises in the Starter Kit. Then you’ll possess 100% clarity about how to write an engineering statement of purpose that will make a huge impact on your applications.
As always, I’m grateful to applicants like Haneesh who allow me to publish their essays so we can all learn from their success. Feel free to explore this blog to find sample essays from other students who earned top admissions. But more than anything, just do me this one favor:
Do NOT write an autobiography essay.
