What did you just type about me, keyboard warrior? Heads up: Im a 17-year-old Indian coder who eats algorithmic puzzles for breakfast, snacks on Codeforces and LeetCode during study breaks, and treats tricky edge cases like collectible achievements; I dont just solve problems, I redesign them, reframe them, and then implement solutions so clean your future self will thank you in comments — recursion, DP, graphs, greedy, segment trees, bitmasks — call them what you want, I call them my warmup routine, and Ill turn a messy problem statement into a polished, tested solution that runs in less time than it takes you to rage-quit the page. Im a competitive-programming junkie and a systems thinker in the same brain: I grind contest rounds and then write editorials because sharing makes the win sweeter, I practice time limits like theyre a rhythm and binary search like its a mantra, and I chase that satisfying "Accepted" green until it becomes a streak — but Im also the kind of dev who cares about production: full-stack web dev isnt a distraction, its a toolkit. I build responsive UIs that dont choke on mobile, craft accessible interactions, wire up RESTful and GraphQL APIs that behave under load, and ship microservices that communicate politely; I write clean CSS (no Frankenstein frameworks unless the UX asks for it), React components that are actually reusable, Node backends that handle auth correctly, and databases — relational or NoSQL — that dont spontaneously implode because I thought a VARCHAR(255) was a personality trait. Generative AI and LLMs are my playground and my obsession: prompt engineering is a craft I tinker with daily, I curate tiny datasets and fine-tune checkpoints, I evaluate perplexity and human-facing outputs until the hallucinations learn manners, and I prototype pipelines that turn messy user prompts into reliable drafts, snippets of code, docs, or creative output; Im comfortable with PyTorch and TensorFlow, I sandbox experiments, I log metrics, I push weights to checkpoints and Ill happily demo a tuned model that writes commit messages better than most interns. Im into prompt chaining and retrieval-augmented approaches, building small RAG systems to give LLMs long-term context without catastrophic forgetting; Ill engineer a prompt template, back it up with a small vector DB, and watch it answer questions like it has a short-term memory that actually works. Security is serious play for me: OSINT and reconnaissance are part research, part art — I find breadcrumbs, map attack surfaces, enumerate subdomains, and build threat models that actually illuminate weak spots; penetration testing and ethical hacking are hobbies and responsibilities — I practice vulnerability discovery in controlled environments, run CTFs for sport, reverse-engineer binaries to understand real attack vectors, fuzz protocols until the weirdness pops, and responsibly disclose what I find. I use Burp, do safe sandboxed exploitation, write PoCs that prove impact without breaking the glass, and make remediation guides that dont sound like legalese; bug bounties are a cheeky side-quest where I level up persistence, and my reports are tidy enough that maintainers sometimes send thanks (and sometimes a bounty, which is neat). I automate everything: CI/CD pipelines that test, lint, and deploy; containers that dont explode; Dockerfiles that actually cache layers sensibly; local dev environments that mirror prod enough to be useful; and scripts that orchestrate repetitive tasks so I can focus on the creative bits. Im comfortable spinning up a small AWS or DigitalOcean sandbox, running a Kubernetes pod for fun, or just keeping things delightfully simple with a compose file and a sensible Makefile. I write unit tests, integration tests, and honestly a few too many comments because good code is readable code — TDD where it helps, pragmatic tests where it doesnt — and I love a well-written README more than some people love espresso. Hardware and low-level tinkering peek in sometimes: I mess with firmware in controlled settings, experiment with IoT devices to understand attack surfaces and build cool projects, and Ill solder a sensor to a microcontroller if it helps me prototype an idea that needs the physical world. Collaboration is in my blood: I contribute to open source, I pair-program with folks who want to get better, I mentor younger coders because teaching clarifies my own understanding, and I livestream problem-solving sessions occasionally — messy, human, useful — so others can watch the thought process, the wrong turns, and the eventual epiphany. I blog about wins and bugs alike: long-form writeups that document how I solved a nasty DP trick, how I pried open a weird exploit in a controlled lab, or how I tuned a tiny LLM to behave less like a drama queen and more like a helpful assistant. I treat every real project like a mini product: user stories, MVP, iterate, instrument, measure, and then polish; shipping matters more than cleverness alone, so I obsess over perf, latency, and graceful degradation, and Ill profile your slow endpoint and hand back a PR that shaves milliseconds with surgical precision. Im playful with my craft: I write cheeky scripts that annotate console logs with memes, I leave gentle easter eggs in UIs that reward curious users, and I make tiny offline tools that automate mundane workflows for my friends. But Im also responsible — when I find a security issue, I disclose it ethically; when a model misbehaves, I log it and iterate; when a teammate needs help, I explain without condescension. My stack is eclectic: C++ for speed when contests demand it, Python for everything experimental and ML-related, JavaScript/TypeScript for the front-end glue, and SQL when data needs relationships; I know my algorithms cold, but I also know when to trade asymptotic purity for pragmatic, readable code that ships. I collect wins in many forms: accepted submissions, clean PRs, respectfully disclosed vulnerabilities, mini-LLMs that actually help writers, slick landing pages that convert, and blog posts that people bookmark. Im loud when its fun, methodical when it matters, and annoyingly elegant in how I approach problems; keep trash-talking if you must — Ill answer with a well-structured function, a polite pull request, a documented fix, and maybe a blog post titled "How I turned your trash talk into a tutorial." Game on: bring your hardest edge case, your weirdest exploit, your cruftiest legacy codebase, or your vaguest prompt; Ill refactor it into something beautiful, write tests so it behaves predictably, harden it where needed, and leave the repository better than I found it — and Ill do it with a grin, a log file full of clever comments, and enough coffee to politely power a small data center.