Platform / oqueo.io

The infrastructure layer for
local AI.

oQueo builds the infrastructure layer between local AI runtimes and real workflows - portable, self-managed, independent of any single cloud provider.

Visit oqueo.io ↗ ← oQueo.com
// platform

A platform built around real constraints

Not another cloud-dependent AI wrapper. Infrastructure that runs where you run it.

🤖
AI Agent Layer
Local-first No vendor lock-in Configurable per workflow

Autonomous agents designed for specific workflows. Each agent is self-contained, configurable, and runs against your local AI runtime.

🔌
Extensibility
Plugin slot pattern Composable design No core modifications

A plugin slot architecture allows new capabilities to be added without touching core infrastructure.

🏠
Self-Hosted Infrastructure
DSGVO-Hosting GitHub Actions Cloudflare

DSGVO-compliant hosting on dedicated D-A-CH servers, deployment pipelines, and CDN - all managed, all controlled.

// agents

Twelve agents. Six live.

Six deployed and active, two in active development, four in the queue.

oqueo.io - agents // status
$ agents list --all
● quentin-ai active
● brandguide-ai active
● botshield-ai active
● blocker-ai active
● quest-ai active
● formguard-ai active
● billpilot-ai active
● dom-guard-ai building
● seo-ai building
○ knowledge-ai planned
○ automation planned
○ mcp-bridge planned
○ cloud-ops planned
13 agents - 7 active - 2 building - 4 planned
▊
// infrastructure

The real stack

Nine systems. Each chosen deliberately.

DSGVO-Hosting
Dedicated D-A-CH Servers

Dedicated servers in the D-A-CH region. Server hosting and storage stay in the EU. Further cloud services run in EU regions or under signed DPAs.

Supabase
Database & Auth

Postgres-based BaaS for structured data and authentication. Open source, self-hostable, production-grade.

GitHub
Version Control & CI/CD

Every deployment flows through GitHub Actions. Source of truth for code, config, and pipelines. GDPR-compliant via DPA.

Render
Backend Hosting

Lightweight backend hosting for FastAPI services in EU region Frankfurt. Deploys directly from GitHub on push.

Cloudflare
DNS, CDN & Tunnel

Handles DNS, proxying, and the Cloudflare Tunnel for local AI runtimes without opening ports. Global CDN - GDPR-compliant via DPA.

Upstash
Serverless Redis

Serverless Redis in EU region (Frankfurt) for caching, queues, and rate limiting. Pay-per-request, zero cold starts.

Claude & OpenAI API
Cloud AI Inference

Claude API (Anthropic) and OpenAI API as configurable cloud AI providers alongside local runtimes. Provider-agnostic by design. GDPR-compliant via DPA.

Hoppscotch
API Development

Open-source API client for testing and documenting all internal and external API endpoints. Self-hosted.

UptimeRobot
Uptime Monitoring

External monitoring for all public endpoints. Alerts on any downtime before users notice.

// philosophy

Four operating principles

Every architectural decision follows these constraints - not aspirational goals, but operational requirements.

🖥️
Local AI First

Every agent runs against a local AI runtime - Ollama, LocalAI, or LM Studio - via Cloudflare Tunnel. No cloud API keys required at runtime.

📦
Portable by Design

The entire stack runs on shared hosting and a small Render instance. No Kubernetes. No AWS. No mandatory cloud dependency.

🔷
Modular Architecture

Each component is a discrete unit. The plugin slot pattern means new capabilities compose without touching existing modules.

🔗
Open Integration

MCP (Model Context Protocol) is the target integration standard - connecting AI agents to real-world data and tools.