Skip to main content
AI Studio is the primary surface for authoring workflows on HASP. A workflow is a triggered, contract-defined automation that may or may not expose a user interface. Workflows with interfaces (a patient intake form that routes to a clinician dashboard) and workflows without interfaces (sync a HubSpot contact to Stripe and post to Slack when a new deal closes) are the same primitive — authored the same way, executed on the same runtime, governed by the same compliance substrate.

​
The authoring model

AI Studio uses chat-first authoring. You describe what you want; the system compiles a workflow contract; the runtime executes it. There is no drag-and-drop canvas, no component palette, no manual graph editing. Visual surfaces (live app preview, trace view, contract view) exist for inspection — they are never editing surfaces. This design is intentional. It is structurally aligned with the agentic-AI future where humans delegate intent and agents execute under explicit constraints. A chat-first interface trains the right habit: describe outcomes, not plumbing.

​
Core concepts

ConceptWhat it is
ProjectThe authoring container. Holds apps, workflows, entities, integrations, webhooks, and settings.
AppA user-facing UI surface within a project, with an explicit audience (public, authenticated, role-gated, agent-callable, internal).
WorkflowA triggered automation — one or more triggers, a step graph, optional UI surface. The runtime primitive.
EntityA shared data schema used by multiple workflows and apps.
IntegrationAn outbound connection to an external system (Stripe, HubSpot, EHR, Twilio, etc.).
WebhookAn inbound HTTP endpoint. Workflows listen to specific webhooks as triggers.
SettingA named configuration value referenced by name from workflow contracts. Change once; propagates everywhere.

​
Create your first project

  1. Go to app.usehasp.com and sign in.
  2. Click New Project and give it a name (e.g., Patient Intake v1).
  3. The project opens in the project overview — the center pane showing your project’s health, recent activity, and an empty resource tree.
  4. Type in the chat panel to start: “I need a patient intake form that collects name, date of birth, and chief complaint, then notifies the on-call clinician if risk is flagged high.”
  5. The system responds with a compiled workflow contract, creates the intake form app, and wires the notification step. Review the contract view on the right.
  6. Click Test to run the workflow against sandbox data without touching production.
  7. Click Release when ready to publish to your live environment.

​
What the system builds

For the example above, AI Studio compiles:
  • An app (patient intake form, audience: Public) at your-org.usehasp.run/intake
  • A workflow triggered by form submission, with steps: receive payload → run PHI-aware risk classifier → conditional branch → post notification
  • An entity (patient) holding the intake record
  • A setting (on_call_clinician_email) referenced by the notification step
All of this is expressed in an explicit, human-readable contract stored in the project. You can inspect it in the contract view, trace any run through the trace view, and replay historical runs for debugging.

​
Next steps

Project model

Understand the six project member types and how they relate.

Apps authoring

Audience taxonomy, app URLs, and how apps participate in workflows.

Workflow authoring

Trigger taxonomy, deterministic and agentic workflow shapes, and multi-trigger.

Examples

Four end-to-end walkthroughs: intake, care coordination, compliance scan, referral routing.