Create SuperiorFood Formulations WithData-Driven Simulation
Bring us your ingredient. Backbond simulates it across thousands of customer formulations and hands you the ones where it already holds.
Trial and error, balancing many conflicting constraints
The project
A customer wants your ingredient in their product.
The targets
Target properties are defined, and they conflict. Fixing one ruins another.
The search
Hundreds of thousands of potential formulations. With expensive trial and error, only a few extremes can be tested.
The risk
Find that formulation and it still has to clear three more. Any one can end the program.
Backbond co-optimizes for all the targets simultaneously
- Solubility
- Define a Number
- Gelation
- Define a Number
- Heat stability
- Define a Number
- Mouthfeel
- Define a Number
- Taste
- Define a Number
Define
The application and the targets it must hit, plus what the engine may move (pH, ionic strength, protein loading) and what it may not (cost per kg, the customer’s process line).
Simulate
Backbond builds thousands of digital twins and evaluates every one against the targets.
Rank
Ranked by the best balance across every target, and flagged by how confident Backbond is in each.
Test
You go to the bench with the few worth making. By default, these satisfy all targets and constraints.
How Backbond simulates a formulation candidate
Four levels of physics decide what an ingredient does inside the customer's product, and Backbond reasons about how all four interplay to set every spec.
A few example use cases
Every one is the same shape: a handful of targets that fight each other, and one formulation that has to hit them all.
01High-fibre products
A shake must carry 12 g of fibre and not turn gritty. Find the fibre blend, particle size and hydration route that achieve the loading, while keeping mouthfeel, viscosity and cost.
02GLP-1 companion nutrition
A meal replacement must deliver a full day's protein and micronutrients at a GLP-1 user's reduced intake. Find the protein blend, premix and serving format that achieve the profile, while keeping satiety, taste and cost.
03Egg replacement
A meringue must foam without egg. Find the protein, pH and shear that achieve overrun, while keeping set, taste and cost.
04Protein fortification
A ready-to-drink must carry 30 g of protein and still pour. Find the protein blend, ionic strength and stabiliser that achieve the loading, while keeping viscosity, heat stability and cost.
05Sugar reduction
A drink must lose a third of its sugar and keep its body. Find the sweetener, bulking agent and acid balance that achieve the sweetness, while keeping mouthfeel, shelf life and cost.
06Clean label maintenance
A sauce must drop its gums and keep its cling. Find the starch, protein and shear profile that achieve the viscosity, while keeping appearance, freeze-thaw stability and cost.
07Heat stability
A plant-protein drink must survive UHT without gelling. Find the protein grade, ionic strength and stabiliser that achieve heat stability, while keeping viscosity, colour and cost.
08Gut-health actives
A yoghurt must still hold live cultures at end of shelf life. Find the matrix, water activity and coating that achieve the count, while keeping texture, taste and cost.
09Supply shocks
A recipe must survive a doubled ingredient price without a new plant trial. Find the ingredient list, loading and process that achieve the same taste and texture, while keeping shelf life, label and cost.
10Crop and lot variability
A formulation must hold spec across harvests and suppliers. Find the specification window, blend and buffer that achieve a constant product, while keeping yield, lead time and cost.