Evidence based nutrition®
Industry proven, peer reviewed and certified by ISO.
Orugen is pleased to announce the release of Maxwell AI V1.3. Maxwell Ai is an innovative cloud-based software solution that operates with no human bias, ensuring impartial model training. With its simultaneous training capability, Maxwell Ai can efficiently train various types of models autonomously, including conventional, shallow, and deep neural net models, with or without spectra.
Maxwell v1.3
FAQs
Can I use any NIR?
Maxwell is platform agnostic and can use data from BlueSun, Foss, Bruker and handhelds.
How do I access Maxwell Ai?
Contact Fermentrics for a demo and their team will explain the process and set up an account. Developed models can be accessed through the datalake or directly via API.
What if my data isn’t perfect?
Maxwell has the ability to understand patterns in spectra that is damaged (noisy, sensor issues or data acquisition issues).
How much data do I need?
The team will work with you to address any data requirements for your task. Maxwell has the ability to isolate deficiencies in model(s), how much data is required to fix them, and what type of data.
How do I explore a relationship between data (ex: TMR and Manure)?
Maxwell has the ability to build hybrid models (associative effect) by selecting components from two different spectra and synthesizing synthetic spectra to derive an answer.
How fast is training?
Once compiled, training can take as little as 2 hours.
Recent work
Digestibility & GHG (Dairy | Beef): Built the linkage between NIR, Digestibility and GHG measurements.
Diet Location: Built the linkages between diet and manure in the 11 diet regions of North America. Diet or Manure could explain each other and the probability the diet came from 1 of 11 regions in North America.
Equine DIgestibiity (Total Tract): Built the linkage between NIR, Digestibility ,and GHG measurements.