What InnerOS owns
The core concepts: persistent memory, agent registry, task state, evidence, tool permissions, model routing, file/document context, runbooks, operational messages and rollback-aware execution.
Operating Layer
InnerOS is the operating system for governed AI operations.
InnerOS is not a product card or a chatbot. It is the orchestration layer that connects memory, agents, model routing, tasks, tools, events, evidence, devices, servers, cloud providers and human gates into one operating fabric.
Architecture Diagram
From conversation to governed action.
01Human / Voice / UI
Operators ask, approve and supervise.
02RalphiIA
Persistent operational intelligence.
03InnerOS
Memory, agents, models, tools, tasks, guards and devices.
04Local + Cloud Compute
Local-first when control matters; cloud when capability matters.
05Business Systems
Quotes, workforce, field service, security, documents and evidence.
How it coordinates
ChatGPT, Codex, Cursor, Open WebUI, local agents and operational services read a shared live state through MCP and Mongo-backed coordination instead of relying on one fragile conversation.
How it decides compute
Task classes such as coding, heavy reasoning, browser review, tests and fallback can be routed to the right node or provider with explicit evidence.
How it stays governed
Sensitive operations use approval gates, status labels, evidence requirements and public/private boundaries. A public page can explain the architecture without exposing private execution.
Why it matters
InnerOS makes AI useful beyond a chat window: it can remember, reason, coordinate, act through tools and prove what happened.
Server Fabric
Two local nodes, one coordination layer.
The public site describes roles without exposing private IPs or credentials.
Primary Intel Node
Coordination services, browser/build/test work, fallback execution, MCP access, Open WebUI integration and daily operational services.
AMD Local AI Node
Local-first AI capacity for coding, heavy reasoning and private model experiments, with vLLM/ROCm direction and rollback-aware model registry.
RalphiIA MCP
Shared bridge for ChatGPT, Codex, Cursor, Open WebUI and agents. It exposes tools, state, runbooks, documents and operational actions under governance.
Cloud + Edge
Cloud Run, Cloudflare, Google Cloud, Alibaba Cloud and external model providers extend reach when public access, scale or advanced models are needed.
AMDIntelOpenAIGoogle CloudGeminiQwenClaudeDockerMongoDBGitHubCloudflareSupabaseReactTypeScriptPythonFastAPIMCPWhatsAppNotionZKTecoHikvisionUiPathDevpostlablab.ai