Machine Learning LeapAI Platform
User-oriented & enterprise-level "full stack" machine intelligence platform

Automation
Automated deployment of analysis tools, automation of user processes, and comprehensive technical support for users
Facilitation
Friendly user interface, modular design of industry solutions, reducing the threshold for users to use artificial intelligence

intelligentization
Intelligent automatic modeling, industry solution intelligent solution integration, providing users with an intelligent experience
Product Center
Automated, easy-to-use, intelligent machine intelligence software tools provide users with an efficient data analysis environment.
Machine Intelligence Platform Component
Deep Learning Component
DeepNex
Support of heterogeneous servers
Container-based cloud, multi-tenant, flexible resource allocation
Embed multiple popular open-sourced deep learning frameworks
GPU resource scheduling: sharing one GPU card, multiple GPU cards in multiple machines
Flexible storage solution: traditional NFS, large scale distributed object storage
Support distributed machine learning training and prediction
High fault tolerance, high scalability, and high availability
Provide rich monitoring functions of the platform
Machine Learning Component
HyperMiner
automated machine learning platform
rich solution templates
industry solution optimization/customized
easy-of-use user guide
off-the-shelf machine learning methods can be used easily and without expert knowledge
machine learning workflow customized
machine learning components can be easily customized/developed
real-time training progress monitoring
expert-level analytic results
visualization of analytic results
knowledge Graph
HyperGraph
independent R&D graph processing technologies
real-time responses
visualization and interaction
multi source data integration and graph modeling
underlying complex path queries
geo-temporal knowledge graph analysis
dashboard for data statistics and monitoring
user workspace management
private cloud deployment
distributed and standalone versions
APIs for customization