Preloader image AI, Brain & Cognition

Petuum

Most AI and ML solutions are hand-crafted black boxes. Petuum aims to allow companies to create AI/ML solutions that are correct, fast, scalable, and consume minimal computing resources, by using standardized building blocks and methods that are transparent and repeatable.

Petuum endeavors to provide an Omni-source, Omni-lingual, and Omni-mount platform that serves the full spectrum of Artificial Intelligence and Machine Learning applications.

Our platform can be used to realize a wide spectrum of AI/ML technologies, such as Regression Models, Deep Learning Models, Graphical Models, Kernel and Spectral Methods, Parametric and Nonparametric Bayesian Methods, Tree and Ensemble Methods, using state of the art algorithms based on optimization, Monte Carlo, and matrix & tensor algebra.

Together with our clients and developers, we enable verticals such as healthcare decision-making, autonomous pilot, anomaly detection and risk management, and beyond.

Omni-Source – processes and integrates data in different formats – such as numeric, textual, imagery, tabular, structured or unstructured, static or streaming – from diverse sources like social media, consumer profiles, electronic health records, sensor logs from IoT devices, transaction logs from financial systems, and machine logs from manufacturing equipment

Omni-Lingual – programmable with multiple popular languages such as Python, R, Java, C++, Julia

Omni-Mount – supports different hardware platforms such as datacenters, workstations, laptops, mobile, and embedded devices