Slow Starts, Strong Finishes: How Precog Shaped My Thinking

I (Harsha Vardhan Nemani) first interviewed for Precog at the start of my fourth semester—armed with little more than curiosity, some enthusiasm, and the vague sense that research might be worth exploring. The interview, to put it mildly, didn’t go particularly well. I walked out convinced I’d missed my chance. For months, I assumed that door had quietly closed. Then, unexpectedly, Prof. PK reached out. He had a project that happened to align closely with my personal interest in finance, and just like that, I found myself at Precog.

What followed was not instant productivity. I spent nearly two years doing what could generously be described as “figuring things out”—switching topics, chasing ideas, getting inspired by people at Precog who were always supportive, and slowly gravitating towards computational social science. Through all of this, PK’s support never wavered. He backed me even when there wasn’t much to back. Eventually, things started to click, papers happened, and I began carving out my own direction. But more than any publication, what Precog gave me was the confidence to navigate uncertainty, and the trust that sometimes, a slow start leads exactly where you need to go.

My journey with Precog began with a genuine interest in finance and a lucky alignment—PK had a couple of finance-oriented projects just as I was looking to get involved. But what truly drew me in wasn’t just the topic—it was the lab itself. Precog had a reputation for being intense, driven, and full of people doing work that mattered. Early on, I got to interact with folks like Priyanshul, Arnav, and Mehul—people who weren’t just experienced researchers, but also incredibly open to helping someone like me find their footing. What was supposed to be one exploratory semester quickly turned into two, even though research output during that time was minimal. The finance projects I started with didn’t quite take off, but every week brought valuable lessons: how to read research critically, how to scrape and clean data, how to structure problems, and how to ask better questions. Even when the work didn’t result in papers, it was building something more important—a foundation for thinking, exploring, and eventually contributing.

At Precog, the learning curve was steep, but never solitary. Meetings and WUs were intense—fast-paced discussions, bold ideas, and quick pivots from theory to implementation. But what could’ve felt overwhelming was balanced by the sheer openness of the group. Feedback flowed freely, not just from Prof PK, but from peers working across wildly different domains. Everyone brought something to the table, and it wasn’t unusual for someone working on a seemingly unrelated project to offer just the insight you needed—or occasionally, a dose of Reviewer 2-style criticism that, while brutal, often helped refine experiments long before they faced the real peer review gauntlet.

One of the most liberating parts of the experience was the way PK approached mentorship. He never enforced a top-down vision. Instead, he created space for detours, dead ends, and slow burns. I was given the freedom to explore, even when that meant months of circling ideas with little to show. He was patient in a way few advisors are, always ready to listen and guide, but never rushing the process. That kind of trust helped me stay grounded, even as I shifted away from finance and deeper into computational social science.

By then, the lab’s research interests were changing, but computational social science still had a strong presence. I found myself among the last few leaning heavily into it along with Tejasvi and Rohan, and that came with a certain sense of responsibility. It wasn’t always easy, but it was deeply rewarding. The ecosystem Precog fostered—intense, curious, collaborative—made sure that even when the work was tough or the direction unclear, you never felt like you were doing it alone.

My initial months, well, years, were rough. This was around the start of my 4th year, following a couple of semesters of trying to make finance-related projects work. Despite the personal interest, nothing really clicked. The ideas felt scattered, and I wasn’t producing much. It was frustrating. That’s when PK suggested I consider working with Tejasvi and Prof. Ashwin Rajadesingan. Their work sat at the intersection of elections, political discourse, and computational social science, a territory that was instantly engaging.

Even after shifting domains, things didn’t instantly fall into place. I was still juggling remnants of earlier projects, unsure of where to focus, and feeling like I was making little headway. Eventually, the finance work naturally fell out of the loop, and PK suggested I streamline my efforts, focus solely on the new project with Tejasvi and Prof. Ashwin. Alongside that came his usual advice: build structure into the way you work. He nudged me to maintain a Notion board, track progress consistently, and approach research with systems that made the messiness more manageable. I wasn’t instinctively drawn to that level of organization, but I stuck with it. Over time, I began to see how these habits helped me think more clearly, work more sustainably, and finally start building momentum.

Working with Prof. Ashwin was invigorating. Around the same time, I also began collaborating with Prof. Vinoo Alluri on a couple of projects that brought a completely different interdisciplinary perspective. Things finally felt like they were moving. Our paper with Tejasvi was the first real push, but it faced a couple of tough rejections. Each one stung, but in hindsight, that process of tearing it down and rebuilding it made the work sharper, stronger, and more grounded. Paper that we published with Prof. Ashwin, Framing the Fray: Conflict Framing in Indian Election News Coverage.

Still, it wasn’t easy. There were stretches where progress felt invisible, and it was hard not to question whether any of it would amount to something. Prof. PK’s advice during that time was steady and simple: focus on the research. He gave just enough direction to keep me moving, but left enough space for me to figure things out. That balance proved crucial. Around this time, an opportunity arose to work with Prof. Kiran Garimella on a new project, one that I could shape from the ground up. It was a natural next step, and one that allowed me to take more ownership of my work. PK remained in the background—always supportive, always trusting, offering the kind of guidance that helps you grow without ever feeling like you’re being steered.

Working with Prof. Kiran Garimella began with a project that, on paper, seemed promising. We were excited about the idea and dove in with energy, but as weeks passed, the results didn’t quite land. The direction we were pursuing wasn’t yielding anything substantial. It could’ve easily been discouraging, but what made the experience different was the space to reflect and pivot. In one of those explorations, a related but entirely unexpected idea emerged, something far more compelling, interdisciplinary, and intellectually rich.

That idea quickly took over most of my time. It combined elements from a lot of fields to answer a question that felt both novel and consequential. For the first time, I found myself voluntarily spending hours digging through data, refining hypotheses, and iterating on models, not out of obligation, but out of genuine excitement. The work didn’t just feel research-worthy; it felt meaningful.

Prof. Kiran played a big role in shaping the project once it picked up. His feedback was clear, constructive, and always grounded in both theory and real-world relevance. Our weekly check-ins weren’t just about progress—they helped sharpen ideas, cut through confusion, and maintain steady momentum. His mix of structure and encouragement helped me stay focused and push the work forward with clarity.

By the time I got the opportunity to work with him at Rutgers during my 5-2 semester, much of the core research had already been done. But the collaboration deepened, especially through the writing and revision process. We first submitted the paper to WWW, faced a tough rejection, then overhauled and refined the work before submitting it. Looking back, it remains one of the most fulfilling projects I’ve worked on—not just because of the result, but because of the process, the mentorship, and the resilience it taught me. Paper that we published with Prof. Kiran, Analyzing Patterns and Influence of Advertising in Print Newspapers.

Looking back, Precog has been far more than just a research lab; it has been a place where I learned how to think, how to persevere, and how to grow. What began as a hesitant step into research has become a defining part of who I am, shaped by the trust of mentors and the company of peers. As I move on to the next chapter (working as ML Engineer at Meltwater), I carry with me not only the skills but also the lessons of resilience, curiosity, and collaboration that Precog instilled in me. Farewells are never easy, but I leave with deep gratitude and the confidence that the spirit of this lab will continue to inspire me, wherever I go.