The only NIR-based software in the cloud

Chemometric Brain software is hosted in the cloud, it can be used anywhere by anyone at scale, sharing NIR spectra and analysis with all types of devices or equipment.

Accessibility

The NIR spectra are available anywhere at any time. This is a good solution for users working in multiple locations, because it fosters collaboration. All spectra from different devices consolidate in a single place.

No hardware required

Working in the cloud, no specific hardware is needed to use Chemometric Brain, thus reducing costs in computer equipment and maintenance.

Easy implementation

To start using Chemometric Brain, only an internet connection is needed. The user just needs to log in with username and password and can immediately start working with his own data.

Centralization

Companies can centralize approvals from any production site and ensure the use of a globally replicable food quality system based on traceability and security.

Tailored qualitative and quantitative analysis

With the implementation of Chemometric Brain software-as-a-service for quality control in any food company, 

the customer can have access to tailored qualitative and quantitative models specifically built to meet the companies´ needs.

Qualitative analysis

Certify compliance of batches or samples with respect to the previous batches received. This methodology helps food companies with routine quality control tasks, such as raw material identification, qualification or rejection in accordance with internal defined standard libraries, food fraud detection, redefinition of products, approval of new ingredients and recipe suppliers worldwide, determination of mixture composition and products shelf-life, surveillance of suppliers, assessment of product homogeneity and supplier accreditation.

Quantitative analysis

Determine the composition of products and ingredients, such as moisture, protein, fat, pH, dry matter, viscosity, ash…. or any others.

Qualitative analysis to identify products’ “fingerprints”

Given a set of NIR spectra from good samples for a raw material or final product, Chemometric Brain’s algorithm analyzes new samples and identifies how each of their ‘fingerprints’ fits within a standard library previously created for the same product. The software then projects the information from the new samples onto the qualitative model existing in the database and highlights any variation from the original product.

Machine Learning for prediction

Machine Learning techniques are implemented in all the functionalities. The system is first taught to develop methods to predict characterization or properties. Most of the chemometrics algorithms are based on machine learning. Because of that, the set of “good” samples or the reference values given to the spectra are so important.

AI to identify blends’ composition

AI is deployed in Chemometric Brain to predict the composition of a final product blend. The system learns with thousands of data about ingredients, percentages and their response to the NIR radiation; after that, the system can give a prediction for any spectrum. Even for those cases in which the concentration of ingredients is very low, having enough number of samples in the library, it is possible to perform an accurate prediction of the ingredients’ percentage in the blend.