Hi, I’m Dev. Shubh Mehta, Independent developer and freelancer in Dahisar, Mumbai, India.

Before programming, I was an artist. I drew, sketched and painted, and I loved making something out of nothing: turning a thought into a shape someone else could see.

That part of me never left. The medium changed. Today I work with code, interfaces, algorithms, 3D scenes and AI models, but it is the same act: imagine it, then make it real.

Tony Stark imagining an invention, engineering the parts and wiring intelligence into it. JARVIS as a companion that remembers. EDITH as connected intelligence. Vision as the question of what intelligence means.

I don't want to imitate fiction. I want to understand the real technology behind those ideas and see how far I can take it myself.

JARVIS is my local AI companion. Python, Ollama, local models like Qwen and Llama, embeddings, a persistent memory database and semantic retrieval, so it can keep useful context between conversations.

The goal isn't a chatbot. It's learning how intelligent systems are designed, connected to tools, and made genuinely useful for coding, research and everyday automation.

As an independent developer, and through Mehta & Sons Infotech, I build for real businesses: a real estate platform, an interior design studio, an artist’s portfolio, a fashion store on Shopify, a food brand and a leather goods manufacturer.

Client work taught me how real software gets scoped, reviewed and delivered, and to listen before I write a line. I do it alongside my degree.

SageStock is my fintech project: real-time market data, broker integrations and AI-assisted analysis, on Rust, Axum, PostgreSQL, Redis, WebSockets, server-sent events and Flutter.

TAC, The Analyst Club, is about learning together: analyst-led courses, assessments, learning paths and live sessions. Some parts are built, some are still on the drawing board, and both teach me that architecture matters as much as the idea.

Minecraft AI agents that follow instructions and use tools. A custom Minecraft network with its own modes. Ethical hacking, and how trust fails inside a system. IoT, sensors, gesture and eye-controlled interfaces.

Underneath it all: algorithms, memory, networking, caching, distributed systems. I want to know not only what a company uses, but why it exists and what it costs.

Having ideas is not the same as finishing them. My work now is turning curiosity into execution: pick a problem, break it down, ship a working version, learn from what breaks, improve it.

I want to grow into a strong software engineer and then a machine learning engineer, work with talented people, and eventually build products and companies of my own.