Skip to content

PRODUCTS

TCG DIGITAL > Tcg mcube features

tcg mcube's feature set

Overview of features

Advanced Analytics and Machine Learning capabilities for diverse user personas

  • State-of-the-Art Technology Stack
  • Intuitive, Actionable Visualizations
  • Massively Scalable and Highly Performant
  • Pre-built Use Cases
  • 500+ Statistical Algorithms
  • 20+ Live Installations Globally

BI features

  • Self-service Capabilities
  • More than 22 Chart Groups with 500 Variants
  • Report Bursting
  • High Volume Queries
  • GUI Based Dashboard Creation with Context Menu and Editing
  • Zero Footprint Solution
  • Sophisticated User Authorization Module
  • Search bar: High-performing, Full-featured Text Search with Cross Field Search

More BI features

  • Interactive Report Capabilities
  • Web-based Authoring
  • Document Layout and Cosmetic Control
  • Dashboard/Scoreboards for Key Indicators
  • Report linking
  • Report Development with Ease of Use
  • Ad hoc Query Generator
  • Metadata Management
  • Web-based Authoring
  • Document Layout and Cosmetic Control
  • Graphical Capabilities
  • Dedicated Analytical Business Application Suites
  • Time-Based Scheduled Reporting
  • Versioning and/or Report Archiving
  • Dedicated BI Portal
  • Integration with 3rd Party Portals
  • Reactive Framework
  • Administrative and Data Security
  • User Profile Controls
  • Usage Monitoring
  • Technical/Architectural Specifications
  • Open Application Programming Interfaces (APIs)

Data Integration features

  • Data Store: Multi-node Big Data Store Comes as Part of the Platform
  • Real-time streaming: Kafka(providing a unified, high-throughput, low-latency platform for handling real-time data feeds) Comes out of the box with Configurations Setup to the Kafka Cluster
  • ETL: Workflow Based Advanced ETL Capabilities with Automated Batch Upload Schedules and with a Large Repository of Pre-processing Operations is Provided as Part of the Platform
  • Query Multiple Data Sources
  • Batch Ingestion
  • Multi-Threaded Ingestion
  • Queuing
  • Technical Metadata querying
  • Custom Coding for Transformation in SCALA, Java and Python
  • Data Upload from Front-end
  • Joining Datasets from Front-end

Features of our drag drop based Advance Analytics workflow creator

  • Drag Drop Based Model Creation without the Need for Writing Code
  • Ease of Changing the Algorithms (eg. Change from Random forest to Logistic regression)
  • Option of Choosing H20 and Spark Libraries
  • Tensorflow Supported
  • Ease of Tuning Model Parameters
  • Some Machine Learning Algorithms available on Drag-drop
    • AutoML: Automated Machine Learning
    • Regression Models – Linear Regression, Decision Tree Regression, GBT Regression, Random Forest Regression, H2O Neural Networks, Isotonic Regression, AFT survival Regression, XGBoost
    • Classification Models- Logistic Regression, Decision Tree Classifier, GBT Classifier, Random Forest Classifier, Neural Network, Naïve Bayes, XGBoost
    • Clustering Methods-   K-means Clustering
    • Chi-square Method for Feature Selection and ALS for Recommendation System
    • Model Evaluation for Regression, Binary and Multi-class Classification
  • Custom Transformations Provided Using R, Python, and SQL
  • Custom Evaluators in R and Python
  • Read and Write to Multiple Data Sources
  • Hyperparameter Tuning
  • Reports/ Visualization
  • PMML Import Supported
  • Workflow Sharing with Admin Governing this
  • Advanced setting: Option of Choosing Standalone, Mesos , Yarn for Connecting to Remote Spark Clusters
  • Scheduling Workflows
  • Detailed Documentation of Each Node and Parameter
  • Pre-built Templates which Ship with the Product
  • Export and Import Workflows

Analytics engine and distributed computing features

  • Multi-threaded, High Throughput Processing for Analytics Comes out of the box
  • These Analytics Tools Run as a Part of the Solution: SparkR, PySpark, Spark SQL, Scala, H20, etc
  • The Spark Environment Supports Classic R as well as SparkR, so External R Servers are Not Required to Run R Codes.
  • Cluster Management
  • Advanced Analytics Models
  • Machine Learning Libraries
  • 100x Faster than Hadoop MapReduce and In-memory
  • Java, Scala, Python, R are Supported
  • Spark Streaming for Real-time Streaming