What does Sunthetics do?

Sunthetics is an R&D optimization platform that helps teams design better experiments, learn from small datasets, and decide what to run next so they can reach formulation, process, and scale-up goals faster.

Sunthetics combines Design of Experiments (DoE) with machine learning and Bayesian optimization so users can move from planning experiments to building predictive models to iteratively improving outcomes such as yield, impurity, cost, or other target properties.

What problems does it solve?

Traditional experimentation is often slow and inefficient. Teams may run too many trials, miss important variable interactions, or struggle to decide what to test next.

Sunthetics addresses this by combining Design of Experiments (DoE) and machine learning-driven Campaigns into one user-friendly platform. 

How Sunthetics works

Sunthetics supports two main workflows:

1. Design of Experiments (DoE)

Use DoE when you need a strong starting set of experiments.

With Sunthetics space-filling DoE, users can:

• Define input variables and ranges
• Apply inputconstraints
• Choose the number of experiments to generate
• Create a downloadable, balanced experimental plan that explores the design space efficiently

This helps teams start with better experiments instead of relying on trial and error or traditional full-factorial designs. A training dataset from Sunthetics DoE is the best starting place for your Sunthetics ML campaigns.

2. ML Campaigns

Create a Sunthetics Campaign when you already have data and want to optimize results and determine what to test next to most efficiently achieve your experimental goals. Campaigns are iterative -- newly collected data can be fed into a new Campaign Version and this will help the ML model learn as part of an active learning loop.

Within Sunthetics Campaigns, users can:

• upload historical or newly generated experimental data
• define inputs, outputs, and objectives
• build a predictive model from the dataset
• receive recommended next experiments
• iteratively improve results by adding new data and rerunning the workflow

Campaigns are designed to support continuous learning and optimization.

What users get from Sunthetics

Sunthetics helps teams:

• Plan smarter initial experiments
• Model experimental outcomes
• Identify which variables matter most
• Understand trade-offs between objectives
• Select the next best experiments to collect data for
• Improve performance with fewer total experimental runs

Typical Sunthetics workflow

1. Set up a DoE to generate initial experiments
2. Run experiments and collect results (training dataset)
3. Upload the data into a new SuntheticsML Campaign
4. Configure the campaign (input constraints, output optimization goals, etc.)
5. Review campaign data insights, visualizations, and model information
6. Run the suggested next experiments from your Campaign
7. Add new data results into a new Campaign Version and repeat