# Private, On-Premise & Air-Gapped AI Deployment | iKB

> Compare hosted, private cloud, on-premise and air-gapped iKB deployment: where data is stored and processed, and what needs external access.

Source: https://kbmy.ai/architecture/deployment

[iKB](https://kbmy.ai/) / [Architecture](https://kbmy.ai/architecture) / Deployment options

# Private, On-Premise & Air-Gapped AI Deployment Cloud or on-premise. Choose where iKB runs.

Compare hosted, private cloud, on-premise and air-gapped iKB deployment: where data is stored and processed, and what needs external access.

[Request a demo](https://kbmy.ai/contact?interest=Deployment%20options)

[_Screenshot: iKB model configuration with sample private inference and embedding settings._](https://kbmy.ai/assets/neo2/models.webp) Model configuration iKB / Sample data

How it works

## Three steps to get started.

- 01 ### Choose the environment Start with a hosted pilot or review private cloud and on-premise options against your IT and security requirements.
- 02 ### Check outgoing data Hosted models, external business systems and web research can send data outside your network. Check each endpoint and permitted action.
- 03 ### Plan offline operation Air-gapped means isolated from external networks. Core services and models run locally. Live web research and externally hosted services are unavailable.

Features & controls

## What deployment options includes.

### Hosted

Use a managed environment for an agreed pilot. Test your documents and questions against clear review criteria.

### Private cloud

Use a dedicated environment. Review model endpoints, connections and external services before enabling them.

### On-premise

Run core services on your infrastructure, close to internal systems. Plan local models, workload limits, installation and recovery.

### Air-gapped

Run core knowledge search and answers on local infrastructure and models. Features that need live external services are unavailable.

Architecture and controls depend on your deployment. Test source quality, permissions, privacy, model failures, retention and exports against your requirements.

Continue exploring

## More from architecture.

[Data & retrieval](https://kbmy.ai/architecture/data-retrieval) [Decision Records](https://kbmy.ai/architecture/decision-records)

[View the architecture overview](https://kbmy.ai/architecture)

## Test iKB on your team's questions.

Bring sample sources and clear requirements. Review the answers and controls before rollout.

[Book a demo](https://kbmy.ai/contact)