Deployment Options
DataFuseAI supports multiple deployment models to fit different infrastructure and compliance needs. You can balance convenience, control, and security without changing how the platform works — only where it runs and how it connects to your network changes.
Deployment Models
Select the deployment approach that aligns with your infrastructure and data-governance needs.
Managed Deployment
Most PopularWe host and operate DataFuseAI for you. Application servers, upgrades, patches, and ongoing maintenance are handled securely without exposing your data. Your team focuses on configuring engines, drivers, and connection profiles – not managing infrastructure. This is typically the fastest way to get started and the least operationally heavy option.
Private-hosted (your infrastructure)
ControlledInstall DataFuseAI on your own private servers or cloud environment. We assist with the initial setup, but your team owns day-to-day operations, networking, and maintenance. This model works well when you need tighter infrastructure control but still want vendor support during deployment.
Local on-premise (offline)
Maximum SecurityRun DataFuseAI entirely inside your local internal network, with no external hosting and no internet dependency. Designed for environments where data cannot leave the premises and security policies allow no exceptions.
What Stays Consistent
Regardless of deployment choice, users interact with the same modules and concepts.
Same Product Features
Engines, drivers, connection profiles, queries, pipelines, jobs, dashboards, and tenant/user management behave the same across all deployment models
Multi-Tenant Support
Tenants, users, groups, and permission structures remain consistent, allowing governance practices to carry over unchanged between environments.
Security & Access Controls
Role-based access control, tenant isolation, and secure credential handling apply regardless of hosting model.
Monitoring & Diagnostics
Dashboards, run history, execution logs, and system health indicators are available in every deployment.
Key Benefits
Security-Aligned Hosting
Use managed deployment when speed and simplicity matter, or move on-premise when regulatory or internal policies require full environmental control.
Operational Flexibility
Deployment strategy can evolve as needs change – without retraining users or rewriting workflows.
Data Stays in Your Systems
Across all models, data remains in your own databases, warehouses, and storage. DataFuseAI only securely connects through drivers and profiles; it does not relocate or access your data.
Consistent Experience for Teams
The UI, workflows, and mental model remain stable, reducing friction as teams grow or environments shift.
FAQ
Frequently Asked Questions About Deployment Options
Three. Managed deployment runs the platform for you, so there is no infrastructure to operate. Private-hosted gives a customer or department its own dedicated instance, which suits teams that need separate environments or custom networking. On-premise runs entirely inside your own environment with no external connectivity requirement. The product features are the same in each; what changes is where it runs and who operates it.
Start from your security and operational model rather than from the product. Managed deployment fits when speed matters more than control and there is no requirement for the software to sit inside your perimeter. Private-hosted fits when you need a dedicated environment or particular networking. On-premise is the answer when regulatory or internal policy requires full control of the environment and the data path.
Yes, and that is a common path. Teams adopt the managed model to get working quickly, then migrate to private-hosted or on-premise as internal capabilities and requirements develop. Because the platform is the same in every model, the migration is a change of environment rather than a change of tooling — the pipelines and jobs you built continue to describe the same work.
No. On-premise deployment runs locally with no external connectivity requirement, so the platform reads from your source systems and writes to your destinations entirely inside your own environment. That is the configuration organizations choose when internal security policy or a regulator requires the data path to stay within their perimeter.
No. The same product features are available across deployment models — the same pipelines, transformations, scheduling, query editor, and access controls. The deployment choice determines where the software runs and who maintains the infrastructure beneath it, not which parts of the platform you can use.
In managed deployment, DataFuseAI does — you build data workflows and nothing beneath them needs your attention. In private-hosted, the instance is dedicated to you but still operated for you, which is why it suits departments needing isolation without taking on operations. On-premise puts the infrastructure in your hands, which is the point: full environmental control is what that model exists to provide.
That is configured separately from where the platform is deployed. Execution is directed by the engine profile in use — the DataFuseAI native engine, a Databricks workspace, or a Spark cluster reached through Apache Livy — so an organization can host the platform one way and run its heaviest workloads on compute it already operates.
Not on its own. Compliance frameworks bind your organization rather than a piece of software, so no deployment model produces compliance by itself. What on-premise deployment contributes is control of the environment and the data path, and the platform adds role-based access and audit logs of what ran and who changed it. How those are configured, and the wider programme around them, remain yours.
