lstmbg2.jpg
corex.jpg

Deep Learning


Deep Learning Since 2005

 

SCROLL DOWN

Deep Learning


Deep Learning Since 2005

 

REVOLUTIONARY AI MADE PRACTICAL

With more than a decade of industrial work and research in neural networks Peltarion has a unique experience in a field where academic research has been the norm. Our aim has always been to take the research beyond the lab and into real world systems. In our view cutting edge algorithms are something necessary but not sufficient for making an impact. With a deep understanding and passion for neural networks we make the technology more accessible by making it practical and more useful by making it more reliable.

lstmbg2.jpg

Current Projects


Current Projects
Some of our current work

SCROLL DOWN

Current Projects


Current Projects
Some of our current work

corex.jpg

Platform


SCROLL DOWN

Platform


Platform

The Peltarion Cortex platform, used in all of our projects is a development and production system that adds an industrial grade software engineering layer to the research oriented open source deep learning libraries.

Our Peltarion Synapse / Spine development environment and production server, have a proven track record of reliability and engineering quality. They have been in production use by reliability critical sectors such as the banking industry for years.

Peltarion Cortex builds on that heritage, extending it to the complexities of dealing with very large models trained on specialized hardware while retaining the easy of use, maintainability and documentability. 

The Cortex platform runs on Ubuntu Linux (GPU support) and on most unix based systems (including OSX) on CPU (typically too slow for training, but fast enough for using the system). It has full version control of blocks and models as well as states generated during training and data sets. Models are serialized with their definitions intact and all system configurations are logged and version controlled. This ensures reproducibility as well as traceability. Peltarion Cortex fully supports continuous training with automatic validation and continuous integration of new models.

Peltarion Cortex uses either Torch7 or TensorFlow as a computational backend. At this point we prefer Torch, but that may change as TensorFlow matures.

About Peltarion


We Love Neural Networks

The Peltarion Story

SCROLL DOWN

About Peltarion


We Love Neural Networks

The Peltarion Story

Peltarion was founded in Stockholm, Sweden in 2005 with a passion for all things related to neural networks. The first thing we realized was that the existing tools were not adequate, so we built our own. Synapse 1.0 was released in 2005 and became over the years the tool of choice for people who wanted to use neural networks but did not have the time nor the inclination to write all their code from scratch. Our real passion was to apply neural networks to real world problems and we got to do that  in a very wide variety of fields ranging from building control systems for heavy industry to helping to analyze the migration of dolphins in the Pacific Ocean. As neural networks can be applied to almost any problem where there is data we have accumulated experiences from different sectors such as finance, industry, IT, health, science and many others.

Due to the current extreme demand of deep learning solutions, and a shortage of people to hire we aren't able to take on as many projects as we would like. However, if you have a lot of data and an interesting problem you want to solve with deep learning, please do contact us.

Do you have neural network expertise? Looking for a job? Please do contact us!

Some of our customers