Hi! I’m Aansh Samyani, a third-year undergraduate from the Mechanical Engineering Department at IIT Bombay. While my core background is in mechanical engineering, it was during my second year that I discovered my passion for Artificial Intelligence and Machine Learning. What started as a curiosity soon grew into a deep interest, one that eventually led me to pursue a research internship at the National University of Singapore (NUS), under the guidance of Professor Chaithanya Bandi.
Like many students aspiring to land a research internship, my path was anything but unclear. The process was filled with its fair share of uncertainties, unresponsive emails, and those typical moments of doubt. I spent a lot of time researching potential professors and sending applications, only to be met with silence or polite rejections. It wasn’t until a friend referred me to Professor Bandi that things took a positive turn. His work in AI, particularly focused on improving the reasoning abilities of Large Language Models (LLMs), caught my attention, as it closely aligned with my interests.
After diving into his research, I felt confident this was the perfect opportunity for me, so I reached out. To my excitement, I received a response. Professor Bandi asked me to implement two existing research papers related to his work as part of the selection process. It was a challenging yet enlightening experience, and after some rigorous evaluation and fruitful discussions and presentations, I was thrilled to secure the internship. The project perfectly matched my growing fascination with AI and gave me the chance to work on cutting-edge ideas.
During my time at NUS, my research focused on developing Multi-Agent architectures designed to enhance the reasoning capabilities of LLM systems. The core challenge was creating systems where agents could reason and learn from one another, in a manner that mirrors human-like decision-making processes. Essentially, the goal was to create AI systems that could mimic complex human reasoning through collaboration and learning — an ambitious task, but one that was both exciting and intellectually stimulating.
I spent a lot of my time diving deep into Reinforcement Learning with Human Feedback (RLHF), a technique used to train these models by refining their behavior based on human input. This added an entirely new layer of complexity to the project, as it combined advanced concepts from both Natural Language Processing (NLP) and Reinforcement Learning (RL). My work involved heavy coding in Python, a lot of debugging, and an immense amount of reading. Every day was a blend of trial and error, scouring through research papers, and analyzing the behavior of the models I was working on. There were moments when things didn’t work as expected, and progress felt slow, but it was in those challenging moments that I learned the most.
Unlike a typical corporate internship with fixed working hours, research is more fluid. It’s driven by curiosity and your understanding of the problem, so my schedule was often intense. I found myself deeply immersed in the work, sometimes spending entire days testing new ideas, tweaking parameters, or exploring a new approach. This unstructured yet fulfilling environment taught me the importance of patience and persistence in research.
Life at NUS was an adventure in itself. I’d start my mornings early, with a cup of coffee in hand, as I rushed my way to campus. The regular meetings with Professor Bandi were both motivating and challenging. We’d often meet twice a day to discuss my progress, brainstorm new ideas, and review relevant papers. These discussions were invaluable in shaping the direction of my research and pushing the boundaries of what I thought I could achieve. His feedback helped me not just with the technical aspects, but also in honing my problem-solving skills and developing a more structured research approach.
I also attended a couple of lectures informally, at the NUS School of Computing and NUS Business School, which gave me exposure to interdisciplinary research and allowed me to connect with students from diverse academic backgrounds. Just exploring the campus, connecting with students, and understanding how things worked there was also something that enriched my experience.
Outside of research, living in Singapore was an incredible experience. The country is vibrant, clean, multicultural, and calm in a sense making it an ideal experience if you want to focus. On weekdays, my focus was entirely on my project, but come the weekend, I’d take the opportunity to explore. I loved wandering through Singapore’s streets, visiting Gardens by the Bay, and Universal Studios, and heading out to Sentosa Island at night, which became a weekly ritual. One of the most memorable aspects of my stay was traveling to nearby places with my random flatmates during my free time. Exploring new places, tasting new foods, and soaking in different cultures provided a perfect contrast to the intense work I was doing during the week. It was the ideal way to unwind and recharge for the Monday Blues.
Looking back, my internship at NUS was an incredibly rewarding experience. It wasn’t just the technical knowledge I gained, but also the clarity that I gained of what I wanted in the long run. The work was demanding, requiring me to push myself beyond my comfort zone and explore new ideas. I learned that research is a slow process at times but equally rewarding. There are moments of frustration, where progress feels elusive, but with determination and the right mindset, breakthroughs eventually come.
My advice to anyone considering a research internship abroad is to be patient and trust the process. There will be setbacks, but those challenges are what will make the trip abroad worthwhile. And of course, if you ever have any questions about my experience or the application process, feel free to reach out. I’m more than happy to help!
