Start with the work. Add infrastructure only when it earns its place.
Every Software Tailor AI App already includes an embedded AI Engine and a useful free model. Describe the business use case, hardware and operating boundary; the planner will show whether standalone apps are enough, when AI Server helps, and when a multi-server estate needs AI Gateway.
Nothing is stored while you explore. Prepare the contact hand-off to carry this plan to Contact for this tab session.
02 · Further refinement
Tune the preset.
Personal private assistant · preset values
The result is a decision aid, not an automatic hardware quote. Pilot the actual models and files before procurement.
Apps and models
Begin with the embedded engine.
03 · Operating boundary
What must be central?
04 · Recommended architecture
Standalone AI Apps
Start here
AI Apps1 user or device · AI Suite
Always included
Embedded AI EngineIncluded default model · runs on the device
Every app remains independently useful. A server connection is never required for the local route.
Optional addition
AI Gatewayrouting · health · failover
Central model and inference service
AI Admin Consoleidentity · policy · licences · audit
AI AppsStandalone FreeEmbedded engine and included default modelAI ServerNot neededKeep inference on the deviceAI GatewayNot neededUsed only for a multi-server poolCentral administrationNot neededManage this device locally
Why this path
Recommendation evidence
Hardware fitUse the current endpoint
A modern 16 GB device is a comfortable starting point for the included small model.
Infrastructure planning figureExisting deviceNo central infrastructure selectedSoftware licencesUS$0Free app pathIndicative monthly totalUS$0USD before tax, when calculable
05 · Commercial handoff
Carry only the required quantities into the Store.
No paid licences are required for this starting point.
Only this action saves the configuration, in session storage for the current tab. Contact shows it and prepares an email; nothing is sent until you send that email.
The deployment ladder
Four layers, each with a distinct job.
Always first
Standalone AI App
The embedded AI Engine and included default model provide a useful free path on the device. Pro adds broader model choice tuned to the workload and hardware.
Add when shared
AI Server
Centralise models, compute, provider credentials or an OpenAI-compatible API. The app keeps its embedded engine for local use and fallback.
Add when pooled
AI Gateway
Put one stable endpoint in front of multiple AI Server workers for health-aware routing, failover and scale. Gateway software is included with AI Server.
Add when governed
AI Admin Console
Use central identity, entitlement, policy and audit when organisation-wide control becomes part of the requirement.