Skip to main content

Why DMetaSoul?

DMetaSoul provides a one-stop new infrastructure platform from data source to data value release.

Real-time Data Lake


  • Efficient concurrency and throughput, real-time data stream writing
  • Unified stream-batch table storage, real-time data warehouse, and samples

Multimodal AI


  • Real-time data seamless docking, sparse feature training inference
  • Online ABTest framework, parameter hot loading

One-stop deployment


  • Online and offline cloud native deployment
  • Efficiency SaaS application


Unified stream-batch table storage

Multi-source data real-time lake entry

With only relevant configurations, such as online data sources, the whole database synchronization and real-time entry task can be started. It supports the automatic sensing of new tables and synchronizing table structure changes without human operation and maintenance. The online data is updated to the lake warehouse in real time. The BI reports and large-screen display are seamlessly connected and updated in real time so that key business indicators can be grasped at any time to support business decisions.


Distributed Multimodal AI

Vast scale distributed multimodal training

Billions of discrete features + multimodal model

Big data framework docking, high-performance inference services

Seamless reading of lake warehouse training samples, heterogeneous online resources, elastic balance

Online algorithms apply the microservice framework

Plug-in experimental hot loading, dynamic stratified flow cutting, efficient iteration

Low Code Recommendation SaaS - Industrial recommendation system for rapid deployment

Based on big advanced data and artificial intelligence technology, it can help enterprises overcome code problems, build personalized recommendation systems easily and painlessly without ML expertise, improve click-through rate and retention rates, and boost business growth.


One-stop multi-cloud unified development environment

Data-centric development model

Full link data-driven new development paradigm

Standardized development processes

Industry-standard data intelligent process landing enabling

Multi-cloud unified one-stop environment

Highly integrated, fast opening, multi-cloud unity

Help enterprises fully release business value. Let data intelligence be at everyone's fingertips.

Solve data model

Get through the data and training model connection to achieve batch stream integration.

Solve model training and reasoning

Supports large-scale sparse feature computation and high-performance reasoning with heterogeneous hardware

Solve the model - business landing

Improve the efficiency of online feature development, reduce cost and simplify experimental business logic

A New Infrastructure for One-stop Lakehouse Intelligent