Hello! I'm Ayush, a Computer Science student set to graduate in 2027, with a deep-seated passion for Artificial Intelligence and Machine Learning (AI/ML).
My formal education may not be from an Ivy League, but I've aggressively pursued knowledge through their open courseware, giving me a strong, self-directed grip on my subjects.
My core interests are laser-focused on Natural Language Processing (NLP) and architecting production-level machine learning pipelines. I enjoy both building systems and, importantly, rigorously stress-testing them to understand their limits. I possess hands-on experience in advanced techniques, including fine-tuning Large Language Models like GPT-3.5 and DeepSeek V3 for specialized applications, such as legal system assistance. Currently, I am applying my skills in MLOps and model quantization to develop a personal project: a robust face-anonymizer Python package.
My journey as a builder started early, around 8th grade, inspired by watching shows like Kiteretsu. This drive led to my first real-world experience in high school, where I developed and launched applications. This was my first time dealing with real users, and I grew the user base to a significant 100,000 users a foundational experience in product development and scale. I'm excited to translate this building spirit into solving complex AI challenges.
Away from the code, I enjoy reading Manga; some of my all-time favorites are Vagabond, Berserk, and Vinland Saga. I grew up in Gangtok, Sikkim, and am currently based in Patna, Bihar, carrying that vibrant background into my work. I've been actively working in AI for the last 1.5 years and am incredibly excited about what's next!
I grew up watching shows like Kiteretsu. Kiteretsu was always accompanied by his robo friend Korosuke, and I always wondered what it would be like to have a robot friend that could talk to you, become your friend—partly because I didn't have many human friends back then. I also wondered if we would ever have a real-life Iron Man. These were just thoughts, and I believed they were only possible in science fiction.
Then, one day, I saw Amazon's Alexa. It was then I realized the dream of having a robot friend could actually come true; I was simply amazed by the technology. I logged onto the hardly working computer my dad had bought for me and spent a few hours on the Alexa skills website. I developed some silly, pretty basic apps you just had to type the prompts and the desired results.
That was my clear calling to get into Artificial Intelligence. That incident might have been small, like a spark, but the fire it started was not small. It was mighty, and it is still driving me today.
In high school, I actually published some Android apps that really took off. They were basic utility apps, like Previous Year Questions for different exams, and they were super in demand back then. I put them up, and in about a year, I hit over a million downloads and nearly 100,000 active users! Making some money from ads felt pretty cool you feel pretty rich having that much during high school.
However, the Play Store ended up suspending my apps due to some policy violations. I tried appealing, but it was no use. That's when I really felt how tough the app market can be. It seemed like you couldn't really innovate unless you had a ton of capital to advertise; otherwise, your app would just get lost in the sea of thousands.
So, I decided to part ways with full-time app development. I still build some apps as personal side projects, and they even have a few thousand users, but my main focus shifted. That whole experience, though, taught me a lot about dealing with real users and scaling things lessons I definitely bring with me to AI.