Architecture
AI Studio provides an integrated platform that covers the entire lifecycle of AI model development — from data preparation and fine-tuning to testing, deployment, and management. The platform is designed to help developers, researchers, and enterprises efficiently build, optimize, and operate AI models at scale.
Components
The platform is built around five main components:
Model Hub
Central repository for storing, versioning, and deploying models. It ensures consistency and accessibility across teams and environments.
Model Fine-tuning
A managed service that enables users to train or adapt existing pretrained models to their specific datasets. Supports scalable distributed training.
Model Testing
Provides tools and environments to validate model performance and compare results across model versions before deployment.
Data Hub
Secure and scalable data management service. Handles dataset upload, organization, and linkage with fine-tuning and testing jobs.
User Token
Identity and access management system. Used for authentication, permission control, and API integrations.
How Components Work Together
Users upload and manage datasets in Data Hub.
They fine-tune models using Model Fine-tuning, referencing datasets from Data Hub.
Fine-tuned models are stored, versioned, and deployed via Model Hub.
Performance is validated through Model Testing.
Access and automation are managed securely with User Tokens.
This modular yet interconnected architecture helps you move seamlessly from raw data to production-grade AI models within one unified environment.
Last updated