How Formul.io
Started

I'm Eugene. Software architect by profession — I build platforms, infrastructure, and edge AI systems for innovation startups. Chocolate maker since 2019.

The two careers don't overlap as much as you'd think. Until they did.

COVID, 2020. I was at home like everyone else. As a technical person, there's only so much screen time before something has to change. I decided to learn chocolate.

Started simple: tempering, basic moulded bonbons. Then truffles. Then ganache fillings, caramels, pâte de fruit. The deeper I went, the more confectionery turned out to be — texture science, emulsion stability, sugar crystallization, water activity. Things I'd never thought about before.

Every recipe I tried from the internet had the same issue. They were written for ingredients I couldn't easily buy — Valrhona Guanaja 70%, specific French butters, glucose syrups with particular DE ratings. So I substituted with what was available locally.

And substitutions broke things.

There's no simple "is it OK or not" answer for any of this. Every recipe needed to be recalculated, and recalculating by hand was slow, error-prone, and discouraging.

When I started developing my own recipes — my own flavors, my own combinations — the problem multiplied. Every adjustment to balance taste required recalculating the whole composition. The pace of iteration was painfully slow. This was the biggest motivator to dig deeper.

When I learn something, I go in deep. That's how I'm wired — figure out the principles, check the boundaries, understand the trade-offs, then build something useful on top. Surface-level understanding isn't satisfying.

So I started reading. Food science literature isn't like web development blogs. It's peer-reviewed papers, food technology textbooks, and equations without obvious physical intuition.

The math was tedious. Some of it dry. But it worked — once I understood how aw behaves when you replace one sugar with another, ingredient substitution stopped being a gamble.

An aw meter costs €1,500–€3,000. For a hobby, that's not a tool, that's a statement.

The math gives me what the instrument would, within a known margin of error.

Before building anything, I checked what already existed. As an architect, you don't reinvent — you find what's there and use it.

I started with ganache, because that's where I started in chocolate. The first calculator was for me — to solve my own substitution problem. Then I expanded: pâte de fruit, caramel, dragée, mousse, ice cream, nougat — everything I actually use in bonbons or as standalone confections.

Each calculator is built around the actual physical chemistry of its product. Ganache uses different math than caramel. Caramel uses different math than dragée. There's no universal formula that fits all confectionery — only specialized ones.

I still make chocolate every week. Bonbons, truffles, caramels — same things I started with. I still use Formul.io for every new recipe I develop. The tool grows with my practice; my practice tests the tool.

Beyond making, I'm focused on sharing the knowledge. Writing the science articles in the knowledge base. Helping people who are stuck on the same problems I was stuck on five years ago. Making the math accessible to people who don't have time to read peer-reviewed papers but still want to understand what they're doing in the kitchen.