Machine Learning

Custom design of neural nets in C++

Seamlessly interleave machine learning with your business analytics implemented in C++ and obtain code that is several times faster than Tensorflow on a CPU!

With MatLogica, both training and inference show great speed-up on CPUs.

For problems requiring up to 1000 inputs, AADC by far outperforms tools like Python, Tensorflow, and others. Python-based tools are starting to catch up for very large problems (such as computer vision and linguistics) where we are limited by the memory bandwidth. See the results of the ADBench benchmark below:

Monte-Carlo based calibration

You can easily calibrate complex multi-asset models relying on Monte-Carlo and avoid using inflexible and difficult to derive analytical approximations.

Recalibrate your models

Use real-time data to recalibrate your models and get an accurate and up-to-date information.

Create custom layers

Efficiently differentiate the custom function definitions as if they were there from the start.