First year, B.Tech AI & ML student blending "vibecoding" speed with deep-dive fundamentals. I use AI agents to prototype fast while mastering the low-level engineering that powers them.

Generative AI, machine learning, natural language processing, embeddings, text-generation models, GPU-aware execution, model inference, Python, Jupyter Notebook prototyping, AI/ML experiments.

Current projects include **Study-buddy**, a Jupyter Notebook application for embeddings and text-generation models featuring GPU-aware execution, and **sciptoria**, a generative AI prototype. These projects leverage Python, Hugging Face/transformers, and open-source ML stacks for model inference and NLP experiments. Additionally, work involves "vibecoding" to rapidly prototype AI agents while exploring low-level engineering, GPU acceleration, and automated embedding pipelines.

Ayush Singh is a student at Geethanjali College of Engineering and Technology, pursuing a Bachelor of Technology in Computer Science. They are open to work and are looking for knowledge sharing with peers, community and friendships, and speaking opportunities. Ayush is open to introductions and prefers to be contacted via email.

AI-integrated web applications, TypeScript front-end development, Next.js and React, full-stack prototyping, LLM integration (Claude, LLaMA), API integration, product-first engineering, UI/UX and CSS animations, film pre-production tools, movie and anime discovery engines, JSON data modeling.

Current projects include **preroll**, an AI-powered film pre-production tool; **velora**, a movie discovery platform built with Next.js, TMDB API, and Claude AI integration; and **anispin**, an anime discovery engine. These projects leverage TypeScript, React, and Node.js, focusing on rapid prototyping and LLM implementation (LLaMA/Claude). Tactical work involves developing versioned JSON data models, PDF export features, and responsive UI/UX using CSS animations and API-driven content.