Skip to main content

Katonic 4.0 (May 2023)

This release provides New features, new UI and enhancements.

New Features and Enhancements:​

  • Enhanced, user-centric Interface: We have improved the user interface to provide a more intuitive and user-friendly experience.

  • Katonic Studio's emphasis on no-code environment with bespoke components: Our platform now supports a no-code environment, allowing users to create custom applications using pre-built components.

  • Integration with distributed environments like Dask and Spark: We have integrated with Dask and Spark, enabling distributed computing capabilities for improved performance and scalability.

  • Custom model deployment feature: Users can now deploy any previously developed model using our custom model deployment feature.

  • GPU integration for model deployment: We have added support for GPU integration, allowing users to deploy models using GPU acceleration for faster inference.

  • Git Integration for model deployment: Our platform now integrates with Git, enabling version control and seamless deployment of models from Git repositories.

  • Revamped automation for monitoring Classification, Regression, Audio, Image, and NLP models: We have enhanced the automation capabilities for monitoring various types of models, providing comprehensive insights into model performance.

  • Deployment of models and apps with auto-scaling feature: Users can now deploy models and apps with auto-scaling capabilities, optimizing resource allocation based on demand.

  • Improved model monitoring mechanisms: We have implemented improvements to the model monitoring mechanisms, enabling better tracking and analysis of model performance.

  • Workspace equipped with GPU integration: The workspace now supports GPU integration, allowing users to perform computationally intensive tasks efficiently.

  • Responsible AI implementation via Explain IT: We have introduced Explain IT, which promotes responsible AI practices by providing explanations for model decisions.

  • Successful benchmarking over CPU/GPU on 100GB data using Nvidia RAPID integration: Our platform has undergone successful benchmarking, demonstrating its performance over CPU and GPU using Nvidia RAPID integration.

  • Resolved Spark integration with Kubernetes pipeline: We have resolved issues related to Spark integration with the Kubernetes pipeline, ensuring smooth collaboration between the two.

  • Integration with Snowflake: Our platform now integrates with Snowflake, expanding data connectivity and enabling advanced analytics and data science workloads.

  • Augmented alert system for Node, Pod, and PVC: We have enhanced the alert system to provide real-time notifications for critical events related to Node, Pod, and PVC.

  • Google Cloud Platform (GCP) integration: Users can now integrate our platform with Google Cloud Platform, leveraging its services and resources for seamless deployments.

  • Integration with Airbyte Connectors: We have integrated our platform with Airbyte Connectors, simplifying data ingestion and integration from various sources.

  • Katonic Visualise - integration with customized Superset: We are excited to introduce "Katonic Visualise," our integration with a customized Superset, empowering advanced data visualization capabilities.

  • Upgraded user interface for File Manager: The user interface of the File Manager has been upgraded, improving usability and efficiency.

  • Revamped user interface for Model Experiment: We have redesigned the user interface for Model Experiment, providing a more seamless and intuitive experience.

  • Updated Accelerator with more use-cases: The Accelerator now includes an expanded set of use-cases, offering more options to accelerate workflows.

  • Kubernetes version upgrades: We have upgraded the Kubernetes version to ensure compatibility and take advantage of new features and improvements.

  • Platform installation script updated for various platform categories: The platform installation script has been updated for different platform categories, including Enterprise (MLOps), deploy, data scientist, Business IQ, workspace, and custom installations.

  • Backup automation integrated with platform installation: We have integrated backup automation into the platform installation process, ensuring data safety and reliability.

  • Integration of GCP: We have added integration with Google Cloud Platform, expanding deployment options and enabling seamless access to GCP services.

  • Customized GPU and deployment node pools to segregate production deployment: Users can now customize GPU and deployment node pools to segregate production deployments, improving resource management and allocation.

  • Connector and Superset integration: Our platform now seamlessly integrates with connectors and Superset, expanding data connectivity and advanced visualization capabilities.

Bug Fixes​

  • Bug fixes and enhanced installation: We have addressed reported bugs and made improvements to the installation process, ensuring a smoother experience.