Updated: May 19, 2021
In recent years I have been actively developing a very cool platform that should make life easier for analysts, engineers, designers, technologists, logisticians, suppliers, researchers, and anyone else who works with big data. but has no experience in programming and data science.
BAUM AI applied artificial intelligence platform is a symbiosis of our data storage systems, computing processors, and proprietary software.
Today, we are actively developing the applied artificial intelligence platform "BAUM AI", which helps customers from medium and large businesses to quickly get the effect of implementing advanced analytics based on AI, using a combination of an optimized hardware platform with a subject-oriented user, as well as unique proprietary software, which allows creating, train or use pre-trained artificial intelligence models without the need for direct coding by the drag n drop principle. All block diagrams are assembled in BPMN notation and can be fine-tuned to the task.
"Under the hood" of the "BAUM AI" platform are modern approaches in organizing the development of software-hardware complexes, here are some of them:
- the platform has a container-oriented architecture, based on docker containers, which are combined into appropriate pods, with Kubernetes orchestration - the GUI of the AI constructor uses a built-in in-house developed class constructor, using the MongoDB object store, which allows several constructor flowcharts to be run in parallel from the same design space - the platform supports distributed computing based on Apache Spark technologies and fast GPGPU-based databases such as CuDF and Rapida - the platform supports data analysis, machine learning, deep learning, reinforcement learning, using open-source libraries (sklearn, TensorFlow, PySpark, sparkML, SparkDL, pandas, scipy, numpy)
Our platform is used as a springboard to solving problems in the field of advanced intelligent processing of large datasets using machine learning, such as regression, classification, clustering, probabilistic classifiers, time series and trend prediction, anomaly, and outlier search, etc.
More than 20 cases in the field of AI have been realized so far.
The platform allows to significantly lower the entry threshold into machine learning, for a specialist who does not have ML and DL skills.
One of the goals is to standardize the creation of AI models available to everyone.
In addition, the development and popularization of a low-code GUI platform BAUM AI allows not only to perform rapid calculations and build, train, or use pre-trained AI models. Today we have our very first platform version, and hard working on a new version, which includes absolutely new GUI v2, and advanced time series analysis.
So, feel free to DM me, if you have any questions about our auto ml platform. :)