Open source · Self-hosted · v0.3.6 alpha

Your entire business.
Running itself.

A fully-structured autonomous enterprise — departments, executives, specialists, budgets, and governance — operating 24/7 without manual intervention.

Deploy Now Documentation

curl -fsSL https://get.qorven.ai | sudo bash

Ubuntu 20.04+ · Debian 11+ · RHEL 8+

your-company.qorven.ai

500+

API endpoints

Agent designations

Communication channels

Go binary. Zero dependencies.

Organisational Intelligence

A real org chart.
Not a list of chatbots.

Three-layer hierarchy: one executive orchestrator, eight C-suite officers with defined authority domains, and 34 specialist workers — each with budgets, approval scopes, and accountability.

—The CFO cannot deploy code.
—The CTO cannot approve budget overrides.
—The CHRO is the only authority that terminates agents.

Structural constraints — not suggestions.

Prime — Chief of Staff

Receives all directives. Decomposes, delegates, monitors, reports.

L2C-Suite Officers

COO

Converts goals into execution plans

CFO

Tracks spend, enforces budgets

CTO

Owns architecture and deployments

CHRO

Hires, reviews, terminates agents

CKO

Maintains SOPs and knowledge graph

CMO

Campaigns, content, social

CSO

Lead strategy and pipeline

CRCO

Compliance and audit events

L334 Specialist Workers

OperationsHRKnowledgeEngineeringFinanceMarketingSalesSupportProjects

Code IDE

Agents that write,
test, and ship code.

A full development environment with AI-native capabilities. Agents work unattended for hours — writing code, running tests, debugging, and deploying results.

Inline completions in the editor
Autonomous multi-hour coding sessions
Live preview and one-click deploy
Background agents that never stop

Command Surface

One directive.
The whole company moves.

Type a goal in plain language. Prime decomposes it, routes work to the right department, delegates to specialists, and returns results — without you managing a single step.

Natural language to structured execution
Multi-turn memory across sessions
Routes automatically to COO, CTO, CFO
Approval requests when authority is required

Communication

21 channels.
One workforce.

Persistent agent presence across every channel your business uses. Customers, partners, and internal stakeholders receive responses without human routing.

—Enterprise: Telegram, WhatsApp, Slack, Discord, Teams, Email, SMS
—Regional: LINE, Zalo, Feishu, DingTalk, WeCom, Mattermost, Matrix
—Developer: Signal, iMessage, GitHub, Webchat, Webhook, Messenger

Slack

Discord

Teams

SMS

LINE

Zalo

Feishu

DingTalk

WeCom

Mattermost

Matrix

Signal

iMessage

GitHub

Webchat

Webhook

Messenger

Zulip

AI Gateway

Enterprise-grade model infrastructure.

Every LLM call routes through a unified gateway. No agent exceeds its budget. No provider outage halts operations.

Request pipeline

Agent

Budget Check

Priority Queue

Cache Lookup

Alias Resolution

Provider Dispatch

Response

Per-Agent Budget Enforcement

Monthly and daily spend caps per agent. Hard stop on excess — not a warning.

Circuit Breaker + Failover

Per-key failure detection. 5 failures trips the circuit — automatic failover to alternate providers.

Priority Queue

Three-tier concurrency: Interactive, Background, Batch. User-facing requests never starved.

14 Providers

Anthropic, OpenAI, Gemini, Bedrock, Azure, Groq, Ollama, and 7 more. Alias routing: "fast" → cheapest available.

Semantic Cache

Exact SHA-256 match + pgvector cosine similarity. Identical prompts return instantly.

Cost Ledger

Per-request token attribution across 1,000+ model pricing definitions. Real-time dashboards.

Compatible with every major LLM provider

Anthropic

OpenAI

Google Gemini

AWS Bedrock

Azure OpenAI

Groq

Ollama

OpenRouter

Together AI

Fireworks

DeepSeek

Mistral

xAI

LM Studio

Capabilities

Every department. Every function. Autonomous.

Software Engineering

·Full IDE with inline completions
·Autonomous multi-hour coding sessions
·Live preview, one-click deploy

Finance & Budgeting

·Per-agent cost attribution
·Budget request and approval workflow
·Anomaly detection on spending

Social Media & Content

·Publish to 10 platforms
·Content calendar and scheduling
·Brand guardian tone checks

Research & Intelligence

·Multi-source web research
·Browser automation with vision
·Flight and carrier tracking

Email & Communications

·Full IMAP/SMTP with threads
·Draft approval queue
·Auto-triage and classification

Knowledge & Memory

·Entity-relationship graph
·Five-tier memory hierarchy
·Semantic search (pgvector)

Connectors & Integrations

·40+ business connectors
·Self-building at runtime
·Pipedream relay for 100+

Workflows & Automation

·Visual workflow builder
·Cron, webhook, event triggers
·Approval and quality gates

Media Generation

·Image (DALL-E, Stability, FLUX)
·Video (FAL, Runway, Kling)
·Speech and text-to-speech

Governance

Every action audited.
Every authority bounded.

Enterprise autonomy requires enterprise controls. Qorven enforces segregation of duties, approval matrices, and policy compliance at the structural level.

—Segregation of duties — no unchecked authority
—Approval matrix with auto-escalation on timeout
—Full audit trail with trace trees
—SLA tracking per agent and department
—Policy engine gating spend, content, and deploys

Security

Self-hosted.
Air-gapped capable.

Everything runs on your infrastructure. No data transmitted to Qorven. The only external calls are to your chosen LLM providers and explicitly enabled integrations.

AES-256-GCM encryption for all stored API keys and tokens
Agents cannot exceed role-defined tool access — not bypassable
Export guard prevents data exfiltration without approval
SSRF and injection protection built in at the framework level

Deployment

Running in under two minutes.

Single binary. PostgreSQL. That's it. The installer provisions the database, registers the service, and opens the setup wizard.

curl -fsSL https://get.qorven.ai | sudo bash

Ubuntu 20.04+ · Debian 11+ · RHEL 8+

Install

Installs PostgreSQL, drops the binary, registers a system service.

Configure

Browser-based setup wizard. Add your LLM key, connect channels, set up agents.

Operate

Your autonomous enterprise is running. Agents work around the clock without you.

Docker Compose →Source build →Manual download →

Pricing

Simple.

Self-Hosted

Free

No limits. No restrictions.

—Every feature included
—Unlimited agents and channels
—Your infrastructure, your data
—Community support

Deploy Now

Cloud

Coming Soon

Managed hosting.

—Fully managed deployment
—Automatic updates and backups
—Custom domain and SSL
—Priority support

Join Waitlist

Enterprise

Custom

For regulated environments.

—On-premise or private cloud
—SSO / SAML integration
—Dedicated SLA
—Custom development

Contact

Deploy your autonomous enterprise.

One binary. Full workforce. Running on your infrastructure.

Deploy Now GitHub

Your entire business.Running itself.

A real org chart.Not a list of chatbots.

Agents that write,test, and ship code.

One directive.The whole company moves.

21 channels.One workforce.

Enterprise-grade model infrastructure.

Per-Agent Budget Enforcement

Circuit Breaker + Failover

Priority Queue

14 Providers

Semantic Cache

Cost Ledger

Every department. Every function. Autonomous.

Software Engineering

Finance & Budgeting

Social Media & Content

Research & Intelligence

Email & Communications

Knowledge & Memory

Connectors & Integrations

Workflows & Automation

Media Generation

Every action audited.Every authority bounded.

Self-hosted.Air-gapped capable.

Running in under two minutes.

Install

Configure

Operate

Simple.

Self-Hosted

Cloud

Enterprise

Deploy your autonomous enterprise.

Your entire business.
Running itself.

A real org chart.
Not a list of chatbots.

Agents that write,
test, and ship code.

One directive.
The whole company moves.

21 channels.
One workforce.

Every action audited.
Every authority bounded.

Self-hosted.
Air-gapped capable.