I'm Aditya. I build products at the edges of disciplines — engineering, finance, philosophy. Doing CS at St Andrews .
At 18, I embarked on a road of quests to transform how people interact with technology.
An integrated thinking environment — where writing, research, and discussion live together. Solo-built. 75+ orchestrated AI models, 114 languages, millions of monthly requests.
A single API endpoint routing to 17 LLM providers. Drop-in OpenAI replacement with PII cloaking, circuit breakers, and full request logging. Born from brainful's AI layer.
Multi-agent market intelligence across equities, crypto, commodities, and currencies. Synthesizes data, social sentiment, and fundamentals into trading signals.
Python library for securing LLM outputs — preventing sensitive data and PII leakage from AI applications.
Intelligent compaction and embeddings designed for high-accuracy retrieval. All on local inference — no data leaves the infrastructure.
Born on the 23rd in India, grew up across seven countries — 🇺🇸 🏴 🇧🇪 🇮🇹 🇱🇺 🇫🇷 🏴. Got hooked on game development and graphics as a kid, then took Harvard's CS50 and it pulled me into the world of the web — the pace, the tooling, the sheer range of what you can build.
In '23 I started brainful from my room in Italy and began an MSci in Computer Science at St Andrews — both in the same year. brainful shaped my foundations in systems design, infrastructure, and AI at scale.
As brainful grew, so did the problems around it. We were sending vast amounts of PII to external LLM providers, so I built llmshield to lock that down. The AI orchestration layer kept expanding until it made sense to separate it into its own service — cailos. The same happened with our embedding infrastructure, which became polyembed. Each product started as a necessity inside brainful and graduated into something standalone.
tidalsight was a different impulse — an exploration into the domains that equally fascinate me: finance, economics, and market intelligence. Multi-agent AI applied to a completely different problem space.
I'm always looking for opportunities to apply what I've learned to new domains and intersections — that's where the most interesting work tends to happen.