Why I Built Glyc — Diabetes and Calculating Glycemic Load One Recipe at a Time
Many people, as they hit middle age or order, begin to fall into the pre-diabetes category. They are told to watch their diet, eat more protein, less sugar, more balanced meals, and often told to try the Mediterranean diet. I was one of those people for a while.
Last fall, however, the scales were tipped just a little too far and I crossed the threshold into the diabetes category. My health was not great, however it was trending the right direction. I had been losing weight and exercising more regularly and my doctor could see that as well. We chose not to prescribe any medicine, but to continue with the better health habits, losing weight, eating healthier meals.
The desire to learn more about how the glycemic index and glycemic load factored into my recipes came about as a result of this. I searched for details and tools to calculate glycemic loads for meals and there was nothing worthwhile or easy to use. Coming from a software engineering background, I decided to help myself and possibly others by doing something about it.
The glycemic load problem
when I started paying more attention to glycemic load rather than just counting carbs I realized the difference matters: two meals can have identical carbohydrate counts but land completely differently depending on the types of carbs, the preparation, the presence of fat and fibre. GL captures more of that nuance than a raw carb count.
The problem is that calculating GL is tedious. You need the GI value for each ingredient, the quantity in grams, and the math to put it together. There are databases and spreadsheets, but nothing that worked with real recipes in real time.
Every recipe I found online gave me a calorie count and a macros breakdown. None of them told me what the glycemic load per serving actually was.
What I wanted to exist
I wanted to be able to paste a URL — or a recipe I had typed out — and get back a clear GL number with a breakdown of which ingredients were driving it.
I wanted transparency, not just a verdict. Not just "this recipe is high GL" but a clear picture showing which ingredients are driving the number and what swaps would bring it down by how much.
That show-the-math visibility is important for two reasons. First, it helps you make better swaps rather than just avoiding entire food groups. Second, it lets you build intuition faster — after you see the breakdown for a hundred recipes, you start to internalize which ingredients move the needle.
Building it
Glyc started as a personal tool. I wanted to solve my own problem first and see if the solution was actually useful in practice before worrying about whether it would be useful to anyone else.
The core of it is an AI extraction layer that reads a recipe — from any URL, in any format — and pulls out the ingredients with their quantities. From there, a calculation engine looks up each ingredient against a GI database and runs the glycemic load math. The results come back with the full breakdown, not just the summary number.
It turned out to be genuinely useful. My dosing decisions got a bit more grounded. I found myself discovering that some meals I had assumed were problematic were actually quite manageable, and some that seemed fine were quietly spiking me.
Who this is for
Glyc is most useful for people actively managing blood sugar — whether that is type 1, type 2, prediabetes, or just an interest in how food affects energy and mood throughout the day.
But it is not only for people with a diagnosis. Glycemic load is relevant to anyone who wants to eat in a way that sustains stable energy levels, supports concentration, and avoids the mid-afternoon crash that follows a high-GL lunch.
The tool is free to use. You do not need an account to extract a recipe and see the breakdown. If you find it useful, I would genuinely like to know — I am still building it, and the most useful feedback I get is from people who are using it to cook real food for real reasons.
If that is you: welcome to Glyc. I hope the math helps.