Year-in-Review
The year my direction finally clicked into place — an ML engineering internship at Kinetic Vision, a first research paper accepted to a NeurIPS mechanistic interpretability workshop, and open-source work that grew into an in-person internship offer from Oumi.
University Honors Program · Year-in-Review Prompt
What specific experiences (honors or otherwise) in the past year have had the most impact on your personal, academic, or professional goals and life trajectory? Which future experiences will further encourage this growth? Please articulate specific academic and/or professional goals for the next year, including how you hope to continue developing your skills, strengths, and areas of professional interest.
Every year I look back convinced the last one set the bar too high to beat, and every year proves me wrong. 2025–2026 was no exception. More than any year before it, this one helped me figure out what I actually want to do with my life, and it laid the foundation for the path I hope to follow through the rest of my academic and professional career.
The first experience that shaped me was my internship at Kinetic Vision as a machine learning engineer. I had been fascinated by AI and machine learning for years, but this was the first time I got to work in the field full-time, with mentorship from genuinely excellent engineers. Those three months taught me more than any course or textbook could have, because I was building real models for real problems from day one, running structural analysis on medical staples one week and working on large-scale data pipelines the next. By the end of the summer, I was certain this was the work I wanted to keep doing.
Around the same time, I started my first AI research project through a summer program called Algoverse. I was paired with a small team of student researchers and a mentor, and given the freedom to propose and run my own idea over ten weeks. I had been drawn to mechanistic interpretability — essentially reverse-engineering models to understand how they actually work — and to AI safety, so I pitched a project on superposition in neural networks. It was approved, and what followed was ten weeks of reading papers, running experiments, and picking apart results. I fell in love with the process. By the end, we had written the work up and submitted it to a mechanistic interpretability workshop at NeurIPS, one of the most prestigious conferences in AI, and it was accepted. That acceptance was one of the biggest turning points of my life. It made me believe I genuinely have a future in AI research, and it opened doors I hadn't expected, including interviews at several AI companies.
The last experience was contributing to open source. Open-source projects are built for a community and by a community, and many companies maintain them in the open so that anyone can contribute. I came across a startup called Oumi whose mission was to democratize AI, and whose main project was a library that makes training AI models far easier. That mission resonated with me, so I decided to contribute, even though I had never worked on a codebase that size before. What started as one small contribution grew into three end-to-end features over the following months. The engineers and co-founders at Oumi noticed, and reached out for an interview. Between that and my research, they offered me an in-person internship at their Seattle office. It was a dream come true, and an offer I couldn't turn down.
The experience I'm looking forward to most is exactly that internship. Getting to work alongside the Oumi team in Seattle this summer feels like the natural next step in everything this year pointed me toward, and I want to make the most of it. Beyond the summer, my biggest goal is to keep growing as a research engineer — getting sharper at turning a research question into experiments I can actually run, and learning the large-scale training and infrastructure side that real machine learning depends on. I also want to go deeper into mechanistic interpretability and AI safety, and hopefully take my research further than a single workshop paper. And as graduation gets closer, I want to start figuring out what comes next, whether that's grad school or a research role, so the momentum from this year keeps building.
This past year has genuinely felt like a dream. Behind every one of these experiences there were plenty of sleepless nights and a lot of hard work, but I am grateful for all of it, and even more excited for what comes next.
