Address.  Minsk, Belarus,  Gorodetskaya 22
Tel. +(375) 29-376-05-63
© 2019 by RadiantBytes. Created with love and care.

Monitoring activities become more and more complex, with a massive amount of rules and exceptions. We offer solutions to create intellectual monitoring systems which can be as much flexible as it's needed because of being educated(or adopted) on real data from the monitored scenario.

 

Machine learning can offer to educate systems to detect objects, people, defects, or scenarios on the video stream. Based on the situation such systems may even react in different ways.

 

As an example: System has detected a high-risk situation it immediately does a call to the responsible manager and gives info about the situation, in case the situation is not critical and detected as and medium severity, the system will just send the report to responsible with alarm and attachments. intellectual systems currently may give a higher accuracy rather than people do.

 

One of the main advantages of such systems is the reaction- it's immediate (sometimes in ms), while a human has to spend some time.

  

Monitoring activities become more and more complex, with a massive amount of rules and exceptions. We offer solutions to create intellectual monitoring systems which can be as much flexible as it's needed because of being educated(or adopted) on real data from the monitored scenario.

 

Machine learning can offer to educate systems to detect objects, people, defects, or scenarios on the video stream. Based on the situation such systems may even react in different ways.

 

As an example: System has detected a high-risk situation it immediately does a call to the responsible manager and gives info about the situation, in case the situation is not critical and detected as and medium severity, the system will just send the report to responsible with alarm and attachments. intellectual systems currently may give a higher accuracy rather than people do.

 

One of the main advantages of such systems is the reaction- it's immediate (sometimes in ms), while a human has to spend some time.

​​

Machine Learning

Neural Networks:
RNN neural networks, YOLO: Real-Time Object Detection adopted model, VGG16 soluton implemented, non-neural network solutions assistance

Tools & Libraries:

Data flow adoptation , OpenCV adaptation, Tensorflow 2.0

Tensorflow, MLFlow, Deeplearning4j

Cloud Platforms and Configuration

Cloud Platforms:

Amazon Web Services, Google Cloud Platform, Microsoft Azure Infrastructure & Configuration

 

Management:

Terraform, AWS CloudFormation; Ansible, Puppet, Chef Languages: Python, Groovy, Bash

Database

 

MS SQL, PostgreSQL, OracleDB, MongoDB, Redis, Redis Sentinel, Neo4J, Cassandra,IBM DB2, Azure Cosmos DB, Cassandra, RavenDB, CouchDB, Azure Storage Tables

Continuous Integration & Delivery

Languages:

Python, Groovy, Bash Build Tools: Maven, Gradle, MSBuild, Nuget, Gulp/Grunt, Webpack CI/CD

Tools:

Jenkins, Bamboo, GitLab Code

 

Quality:

SonarQube

Repository/Artifactory:

JFrog, Nexus, Npm, Nuget

Our Expertise