Le grand modèle inGPT d'Inspur Haiyue, basé sur une base de modèles généraux, utilise une approche « pré-entraînement + ajustement détaillé », et intègre des connaissances et des données commerciales du domaine de l'entreprise et s'adresse aux domaines tels que la finance, les ressources humaines, la collaboration, le développement logiciel, ainsi qu'à des industries telles que la fabrication, l'agroalimentaire, et les services d'eau, et offre des applications intelligentes pour les entreprises, telles que la réponse intelligente aux questions, les documents intelligents, l'analyse de données en langage naturel, la vérification intelligente des contrats, la recherche intelligente, et le développement intelligent en basse code, afin d'autonomiser l'entreprise dans sa montée en gamme vers l'intelligence.
Standard development system
high-performance dynamic modeling
drag-and-drop interface design
and one-stop application development
E2E process driving
graphical process design
all-round process monitoring
and intelligent process simulation
Comprehensive trustworthiness integration
dynamic mapping framework
automated API document generation
and ecological API operation management
Comprehensive device connection and management
mass industrial analysis algorithms with big data
low-code cloud configuration tools
and integrated application integration
Real-time data acquisition
global data government
data value mining
and data asset accumulation
Out-of-the-box intelligent services
high-degree-of-freedom deep learning algorithms
data-driven、self-training
and independent and efficient RPA process automation
Collaborative application delivery concept
integrated development and operation management
and one-click automated application deployment
improving R&D efficiency and product quality
Next-generation ServiceMesh architecture
automatic service registration and discovery
non-perceptual service link tracking
and full-service operation monitoring
Computing resource scheduling management
autoscaling during application operation
data storage and backup
and fast and high-quality containerized delivery capabilities
Cloud native, next-gen microservice architecture: SeviceMesh, and DevOps multi-cloud development and O&M integration
Model-driven architectures, visualized forms, and process design
Front-end Farris UI framework, customized role-based interactive experience
Big data mining, integration, and services enable business analysis and decision.
Sustainable evolutionary large model services based on the Haiyue Large Model empower enterprises to achieve full-service digitalization and intelligent transformation.
Role-based refined permission control ensures high-level security through grading protection and classified protection authentication.
Cloud-edge collaborative architecture with extensive connectivity, real-time data computing, and 3D visualization
The HIP hybrid integration platform offers rapid integration capabilities for different identities, data, applications, and services.
UBML, the low-code core modeling system, is now open-source. This dynamic fosters a UBML low-code open-source community (gitee.com/ubml) and Inspur Haiyue inBuidler open-source community version (ibc.inspures.com). Together, we're building an open, collaborative, secure, and green open-source system.