Skip to main content

Getting Started with the AI Platform

This guide walks you through logging in, completing initial setup, and having your first conversation with Jarvis.

Prerequisites​

  • An active Hub organization (required for SSO login)
  • A Hub user account with the MSP Owner or Technician role
  • The AI Platform subscription assigned to your organization (Free tier is available without billing setup)

Key Concepts​

TermMeaning
TenantYour organization's isolated AI workspace. All data, agents, and usage are scoped to your tenant.
Custom GPTAn AI agent you configure with custom instructions, knowledge, and product scopes.
JarvisThe persistent AI assistant available across all products via Cmd+J (or the HubBar icon).
AI GatewayThe governance layer between your prompts and AI providers — enforces quotas, scrubs PII, and routes to the right model.
StudioThe no-code app builder for creating client-facing portals. Separate add-on billing.
PrimitiveA metered AI service unit: a chat token, a voice minute, a crawled page, etc.
Feature routingThe configured mapping that decides which provider and model handles each product feature (e.g., ticket-suggest → GPT-4.1-mini).

Initial Setup Checklist​

Work through these steps the first time you log in:

  • Log in with Hub SSO
  • Confirm your role and specialties in Jarvis preferences
  • Send your first message in the Playground
  • Create your first Custom GPT (or use a system template)
  • Review your tier and token quota on the Usage page
  • (Optional) Connect Studio to your PSA
  • (Optional) Deploy your first Studio app from a starter template

Logging In​

  1. Navigate to app.theonestudio.app
  2. Click Sign in with Hub
  3. You will be redirected to Microsoft Entra for authentication
  4. After successful authentication, you return to the AI Platform dashboard
  5. Your session cookie (oneai_session) is set on .theoneaiplatform.com — it is valid across all platform services without re-authenticating
ℹ️If you see "Organization not found," your Hub organization has not been provisioned on the AI Platform yet. Contact your Hub owner or support.

Dashboard Overview​

The dashboard is your home base:

  • Quick actions — Start a new conversation, create a GPT, or open Jarvis
  • Recent projects — Your last-accessed conversations, GPTs, and Studio projects
  • Usage summary — Tokens used this month vs. your tier limit
  • PSA integration status — Whether the AI Platform has a live connection to your PSA (relevant for Jarvis context)

If you are already inside another The One product, you can access the AI Platform without opening a new tab:

  1. Click the HubBar at the top of any product
  2. Select AI Platform from the product switcher
  3. The AI Platform opens in your current tab, maintaining your Hub session

Alternatively, press Cmd+J (Mac) or Ctrl+J (Windows) from any The One product to open Jarvis directly without leaving your current product.

Your First Jarvis Interaction (Step by Step)​

Jarvis is the fastest way to experience the platform. Here is how to have your first conversation:

Step 1 — Open Jarvis Press Cmd+J from anywhere in the AI Platform (or any other The One product). The Jarvis panel slides in from the right side of the screen.

Step 2 — Set your context (first time only) Jarvis will ask a few onboarding questions:

  • What is your role? (e.g., Owner, Technician, vCISO)
  • What are your specialties? (e.g., M365, networking, cybersecurity)
  • How many clients do you manage?

These preferences are stored as Jarvis memories and shape every future response.

Step 3 — Ask your first question Type something like:

"Summarize the last 5 open critical tickets for Acme Corp" "Draft a QBR summary for a client with 45 endpoints, 3 critical alerts this month, and 99.8% uptime" "What are the top security risks I should address this quarter?"

Jarvis uses your role context, connected product data (PSA, CRM, Defend), and the AI Gateway to generate an accurate, personalized response.

Step 4 — Continue the conversation The panel maintains conversation history. You can:

  • Ask follow-up questions
  • Request a different format ("make that a bullet list")
  • Ask Jarvis to take an action ("create a ticket for this in PSA")

Step 5 — Close and return Click anywhere outside the panel or press Esc to close Jarvis. Your conversation is saved and resumes where you left off next time.

Using the Playground​

The Playground is the full-featured chat UI for power users:

  1. Click Playground in the left navigation
  2. Select a model from the dropdown (GPT-4.1, GPT-4.1-mini, GPT-4.1-nano, GPT-4o, GPT-4o-mini)
  3. Optionally set a system prompt to shape the assistant's behavior
  4. Type your message and press Enter or click Send
  5. Responses stream in token-by-token

Conversation management:

  • Use New conversation (+ icon) to start fresh
  • Click the conversation title to rename it
  • Use the Pin icon to keep important conversations at the top
  • Drag conversations into Folders to organize by client or project
  • Use Branch to fork a conversation at any point and explore a different direction

API Keys​

If you are integrating the AI Platform into your own tools or scripts:

  1. Click Settings → API Keys in the left nav
  2. Click Create API key
  3. Give the key a descriptive name (e.g., "Zapier integration")
  4. Copy the key immediately — it is only shown once
  5. Use the Authorization: Bearer oneai_your_key_here header in API requests
⚠️API keys are scoped to your tenant. Do not share keys across tenants or commit them to source control. Rotate keys immediately if compromised.

Next Steps​

  • AI Gateway — Understand how your prompts are routed and governed
  • Custom Agents — Build your first agent
  • Usage Quotas — Know your limits before you hit them
  • Studio App Builder — Deploy a client portal in 15 minutes (documentation coming soon)