-
Notifications
You must be signed in to change notification settings - Fork 0
Gen3 Workflows
GT AI OS Release edited this page Jun 18, 2026
·
3 revisions
- Read Getting Started to see which workspaces your role can access.
- Pick a workflow guide below that matches your goal.
- Prepare Datasets and Agents before opening GT Chat.
- For external tools, plan a GT API rollout (requires GT API enabled on your account) instead of manual chat steps.
- Use the ? help shelf (GT Helper) when you want conversational guidance inside the app.

GT AI OS is flexible; most failures come from skipping setup (no dataset, wrong agent, missing paperclip attachment). These workflow articles give repeatable sequences for the use cases operators ask about most—document review, diagram review, and mock assessments—using the same Gen 3 routes you already have in the tenant app.
| Goal | Article |
|---|---|
| Review text documents with retrieval | Document Review |
| Review diagrams and visual artifacts | Diagram Review |
| Run training or evaluation exercises | Mock Assessments |
- Agents — behavior, model choice, default datasets (Agents, Building Agents).
- Datasets — durable source material (Datasets).
- GT Chat — live execution with paperclip dataset attachment (GT Chat).
- GT API — OpenAI-compatible automation (GT API Overview).
The tenant route /workflows (when enabled in your deployment) may expose multi-step automation definitions separate from these instructional guides. Use chat for exploratory work; use workflow automation when your tenant ships repeatable orchestration in that workspace.
- Using Datasets in Chat
- Conversations — review and export completed workflow threads