Four models cover everyone from a proof-of-concept team to a system integrator shipping under its own brand. Same platform, same catalog, same schema — wherever it runs.
Orchestrations move through a controlled lifecycle, validated against the schema before they ever reach your catalog.
Model 01
Deploys from Azure, AWS, or GCP Marketplace directly into your cloud tenant. Ideal for enterprises with existing hyperscaler infrastructure.
Model 02
PHTN.ai hosts the platform for you. Ideal for smaller companies or fast-moving proofs of concept.
Model 03
The full platform, branded with your design system. Ideal for system integrators offering agentic AI under their own brand.
Model 04
The full platform or individual modules, with source code and ongoing updates. Ideal for enterprises who want to own their fork.
Marketplace deployment gives you platform-grade software with SaaS-like operations — without your data ever leaving your subscription.
The platform runs on your Kubernetes — AKS, EKS, GKE, or on-premise — and updates ship as versioned releases you upgrade on your own schedule, with runbooks provided. Internal multi-tenancy lets business units, departments, and subsidiaries share one deployment safely.
Characteristic
How it works
Deployment
You deploy from the marketplace into your own cloud subscription
Infrastructure
Runs on your Kubernetes — cloud or on-premise
Data Residency
All data stays in your VPC; it never leaves your boundary
Updates
Versioned releases; you control upgrade timing
Operations
Your ops team manages infrastructure; PHTN.ai provides charts, migrations, and runbooks
From multi-cloud estates to fully disconnected environments — the deployment model flexes to your regulatory reality.
Published and deployable today from Azure Marketplace, AWS Marketplace, and GCP Marketplace — into your tenant, in your region.
Full platform in your data center, or fully disconnected for defense, government, and critical infrastructure — with self-hosted models and offline artifact transfer eliminating every external dependency.
Split workloads across cloud and on-premise, or run across AWS, Azure, and GCP simultaneously — with cloud-native model backends supported on each.
Time to value: a marketplace deployment is typically up in a day, and a first super agent built from a template takes 6–50 minutes. No intermediary licensing — PHTN.ai does not resell or sub-license third-party products, so you keep direct vendor relationships for models, gateways, and infrastructure.
0 minutes. A live walkthrough of deployment and licensing options.
Book a DemoOr drop us a line at [email protected]