NewConnect Brain to your approved inventory, CRM records, documents and business tools.Explore Live Data Connectors
QualiBot
Log inBook a Demo
Self-hosted enterprise

Run QualiBot in infrastructure you control

QualiBot is self-hosted-first and Docker-native, with an Enterprise overlay for Brain, SAML, audit logs, custom roles, SLA and white-label capabilities. Choose managed cloud or your own infrastructure — the same platform either way.

What's included
Docker and Docker Compose deployment
PostgreSQL with pgvector for Brain knowledge
Redis for cache, jobs and realtime
Storage of your choice: S3, GCS, Azure, MinIO or local
SSO, custom roles, audit logs and white-labeling on the Enterprise overlay
Deployment architecture

A supported, documented deployment stack

The web application, background workers, database and storage layer separate cleanly, so you can size and operate each part independently.

Sizing, high availability and disaster-recovery design are scoped with your team before deployment.

DockerDocker ComposePostgreSQL with pgvectorRedis, including Sentinel-capable configurationActiveStorage: S3, GCS, Azure, MinIO or local storageRails web processesSidekiq background workersEnvironment-based configuration

What self-hosting gives you

Infrastructure location
Private-network access to internal systems when configured
Storage selection
Model endpoint selection through an OpenAI-compatible endpoint
White-label configuration
Backup, monitoring and operational policy

What you own operationally

Server and operating-system security
TLS certificates
Database and Redis operations
Backups and restore testing
Monitoring and capacity planning
Upgrades and incident response
Contractual clarity

The exact responsibility split between your team and ours is agreed contractually before deployment, so ownership of each layer is never ambiguous.

Self-hosted FAQ

Deployment questions

Can QualiBot run on a single server?

Yes, for smaller deployments. Production sizing is based on your actual traffic, file storage, Brain usage and availability requirements.

Is high availability supported?

The application components can be separated and scaled independently. A complete HA architecture, SLA, backup and disaster-recovery design is scoped as part of implementation.

Can we use our own model provider?

Yes, when it exposes a compatible OpenAI-style endpoint and passes quality testing for your use case.

Can this run fully air-gapped?

Self-hosting gives you control over infrastructure location and network policy. Brain, channel providers and connected tools typically need outbound connectivity to function, so full air-gap scope is discussed as part of your architecture review.

Next step

Design the infrastructure and responsibility model before deployment

Bring your infrastructure constraints, and we'll map out the architecture, responsibility split and rollout plan together.