Model providers
Configure provider credentials and model access at the organizational boundary.
Ecosystem
Xpert is open source and plugin-first. Models, tools, integrations, skills, and product interfaces can evolve without turning the platform core into a monolith.
Intelligence
Use hosted APIs, OpenAI-compatible providers, local model runtimes, knowledge bases, semantic models, and business ontology.
Configure provider credentials and model access at the organizational boundary.
Build reusable retrieval pipelines and governed knowledge bases.
Organize enterprise resources, objects, relationships, actions, and policies.
Action
Give Agents the smallest useful action surface, from built-in tools to external business systems.
Use built-ins, code tools, sandboxed actions, and MCP servers.
Connect Feishu, WeCom, DingTalk, GitHub, data sources, and more through plugins.
Shape retries, fallbacks, context, safety checks, and the Agent loop itself.
Build
Create reusable capability packages and deliver them across the Xpert product family.
Contribute runtime providers, tools, views, templates, and packaged Apps.
Integrate Xpert capabilities into external products and workflows.
Package repeatable operating knowledge and Agent structure.
Product family
Build and operate general-purpose enterprise Agent systems.
A coding Agent system for enterprise software delivery.
A data and ontology system for enterprise Agents.
NEXT STEP
Start in XpertAI, follow the technical path in the documentation, or talk with the team about an enterprise operating boundary.