Yogeshwaran
A.
Computer science undergrad who likes systems that can explain their own reasoning, teams that ship in days rather than quarters, and code that's still readable six months later. I build, I tinker, and I'm usually debugging something on the side.
Building agents that show their reasoning, not just their output.
I'm a B.Tech Computer Science student at NIT Trichy (CGPA 8.4), currently working as an AI/Software Engineering Intern, where I design multi-agent systems that debug production incidents across databases, logs, and source code in minutes instead of hours.
Outside of that, I run Bites, a campus food-delivery platform I founded and still operate, and I build research-grade AI systems on the side — a multi-agent stock research committee and an explainable EEG classifier for ADHD detection among them.
I'm drawn to the same problem in every project: how do you make a system's reasoning legible enough that a human will actually trust its conclusion?
A toolkit, not a scoreboard.
No invented percentages here — competence in a language isn't a number I can honestly quantify for you. Grouped by where each tool actually gets used.
Five systems, five different failure modes solved.
Each card opens into the specifics — what broke, what I built, and what it moved.