NanoOrch
A self-hosted, multi-tenant platform for orchestrating AI agents across OpenAI, Anthropic, Gemini, on-prem Ollama, and vLLM (on-prem GPU clusters) , with 3-tier role-based access control, per-workspace resource limits, Docker-isolated task execution, real-time monitoring, approval gates, pipeline/DAG chaining, observability dashboards with utilization threshold alerts, scheduled jobs, two-way Slack/Teams/Google Chat messaging (inbound messages routed to agents, replies posted back to the thread), channel-based delivery for heartbeats/pipelines/jobs/triggers, outbound notifications, cloud integrations (AWS/GCP/Azure/Teams/Slack/Google Chat), DevTools integrations (Jira/GitHub/GitLab), RAGFlow knowledge base support, and a chat UI with @agent mentions.