Alfraido learns from your experimental data and predicts the formulations most likely to hit your targets.
A typical project dashboard. More analysis available.
How it works
Define your ingredients (variables), what you're trying to achieve (objectives. Think strength, stability, cost, etc.), and any limits you can't cross (constraints). Takes about 10 minutes.
Continuous, discrete, or categorical variables all supported
Add results as you run them, one at a time or in bulk via spreadsheet import. Your whole team can contribute, and every entry feeds the model.
CSV bulk import included on all plans
Alfraido builds a machine learning model from your data and suggests the most promising formulas/processes to test next. You can also ask it to predict outcomes for any new combination before you run it.
Multi-objective Pareto front included
Features
Alfraido shows predicted outcomes alongside confidence levels. You can see where the model is reliable and where more data is needed.
When multiple objectives matter, Alfraido maps the trade-offs so you can choose the formulation that fits your actual priorities.
Set limits on inputs and outcomes. Alfraido only recommends experiments that meet your constraints.
Features
Invite your team, and collect data in a collaborative and structured workflow.
Already have results in a spreadsheet? Import them in one go. You can hit the ground running with your historical data.
Every test run is stored and searchable. Nothing gets lost in email threads or personal notebooks.
Where it fits
Balance clinker factor, SCM substitution, and grinding aid dosage against strength and water demand.
Proportion binder, aggregates, and admixtures for strength and workability at lower cost and CO₂.
Hit hiding power, gloss, and flow while balancing pigment loading, VOC, and cost.
Pricing
$ 0 per month
$ 149 per team / month
Custom talk to us
FAQ
Not at all. Alfraido handles all the modelling for you. You supply the experimental data; Alfraido does the maths and surfaces the results in a way that makes sense for a formulator, not a data scientist.
Yes. You can upload a CSV file and map columns to your variables and objectives. If you have years of historical data sitting in a spreadsheet, that's a great starting point for the model.
Alfraido works for any formulation or blend development where you're adjusting ingredient ratios or process parameters to hit target properties — coatings, adhesives, cosmetics, food products, construction materials, specialty chemicals, and more.
The model improves with every data point, but it can start making useful suggestions with as few as 5–10 runs. You can also seed it with historical data to get a head start.
Your formulation data is private to your account and never used to train shared models. We store data securely and never share it with third parties.
Blog
Optimization & R&D
Jul 19, 2026 · 3 min read
A formulator should not have to become a trained data scientist to harness the full power of machine learning. You should not need to know what XGBoost, a neural network, or a random forest is to get results.
Optimization & R&D
Jul 10, 2026 · 4 min read
Most labs that do formulation work end up using LIMS, DOE or ELN software so complicated they require full time data scientist to operate. Alfraido is here to change that.
Optimization & R&D
Jun 14, 2026 · 2 min read
Machine learning assisted materials formulation reads your past experiments results, uses advanced non-parametric modeling to find the underlying trends, then points you when it sees improvement in your objective variables.
No credit card needed. Set up a project, import your data, and see what the model thinks — in less time than it takes to plan your next batch.
Create a free account