Arnav Gupta

I'm Arnav, 20, raising my standard for what it means to be a software engineer every day. I've worked across a wide range of projects and love building things that are practical, meaningful, and feel like they belong in the real world.

Right now I'm building CheckIn, a mood and emotion tracking app that brings together data from health stats, music history, location, and daily activity. It runs on a custom AI engine we built from scratch, trained on our own data and combined with existing LLMs and machine learning frameworks to deliver real-time mood forecasts and emotional insights. The goal is simple: make emotions as understandable and shareable as weather.

Most of my experience comes from CheckIn, the successor to MoodGPT. Building it gave me deep, hands-on exposure to Swift, from UIKit and SwiftUI to Objective-C, CoreML, Metal, HealthKit, and Vision. I've run into and solved many of Swift's sharp edges, from rendering bugs to state sync issues and concurrency timing. Over time I learned to reason through them like both an engineer and an artist. Swift, to me, isn't just syntax; it's a language for building experiences that breathe.

I've also been building Micro Language Models (µLMs), small domain-focused intelligences that can stand alone or connect to larger LLMs. They bring personality, memory, and consistency, the soul layer of AI. I love connecting them into distributed systems, quantized pipelines, and custom RAG layers to form seamless emotional feedback loops.

I've worked with a wide stack of frameworks, tools, and runtimes, and built a few of my own when none existed or when I wanted to push something further. I like connecting distant points on the tech spectrum, from running LLMs on remote servers to blending Blender-style visual logic into Swift, tuning data flows across Elixir/Phoenix, and experimenting with every new framework that challenges the way I think. Curiosity keeps me sharp, and that's the part I enjoy most.