I'm Nick. I am half-British, half-Bulgarian, born in the US and raised in the mighty city of Sheffield, South Yorkshire. Or, as Alex Turner aptly addressed me: “you're not from San Francisco you're from Hunters Bar”.

My first job was at a record shop. I spent two years of salary on vinyl and learnt to DJ at a local working man's club, developing a real passion for listening to, cataloging and sharing music. I still play sets when I can: I recently played at the oldest reggae bar in Japan; a boat party in Hungary; a freezing cold club in the Alps; and parties across the UK and US. If I'm not on my laptop, I'm behind my turntables.

Collecting records taught me something I didn't have a name for until much later: curation. Knowing what to surface, when, and at what intensity. Reading a room. Managing energy across time. This skill is something I utilise when I build.

I studied physical chemistry at Imperial. Why? 1) I wanted to be in London (opportunities), 2) I love science, maths, and coding and computational chemistry, in my head, was the perfect fit, 3) I was rejected by MIT at undergrad and Imperial offered a master's exchange programme. This choice led me to a bunch of interesting internships, placements and research projects (check my resume on the right of the page). I fell in love with digital twins (especially battery modelling — batteries are so complex and so fun to model), did my master's at the coolest electrochemical research group at MIT (the thesis I wrote won the best thesis prize that year and was published in a flagship journal) and so naturally ended up working at an energy storage startup building digital twins and embedded ML models.

Then I spent time at a hedge fund building AI and MCP server architecture for an internal trading system. That's where I started running too many parallel Claude instances at once to get work done, and noticed that the interface for doing this was basically nonexistent. Browser tabs. Terminal sessions. No state. No memory. Context switching.

I realised I wanted to solve this problem. I'm building the visual operating layer for AI agents. The interface that makes directing a team of agents feel obvious rather than chaotic. Most AI tooling is being engineered for capability. Nobody is engineering for feel. That's what we're fixing.

I also love climbing, cooking, reading, playing footy and going to the pub with my friends :) Ok enough blabbering. Peace.