Skip to main content
The API is organized around two creation paths. Both end at the same place: a generated multi-module video course delivered via signed asset URLs. Create a preset once per client company. Reuse it for every course you build for that company. 
1.  POST   /api/v1/presets                          Create preset
2.  POST   /api/v1/presets/{id}/brand-assets        Upload logo + brand colors
3.  POST   /api/v1/presets/{id}/brand/generate      AI derives color palette (async)
4.  GET    /api/v1/presets/{id}/setup-status        Poll until brand = ready
5.  POST   /api/v1/presets/{id}/brand/approve
6.  POST   /api/v1/presets/{id}/slide-style/generate  (async, poll setup-status)
7.  POST   /api/v1/presets/{id}/slide-style/approve
8.  POST   /api/v1/presets/{id}/voice-style/generate  (async, poll setup-status)
9.  POST   /api/v1/presets/{id}/voice-style/approve
10. POST   /api/v1/presets/{id}/questions/start
11. GET    /api/v1/presets/{id}/questions/next        loop: get question
12. POST   /api/v1/presets/{id}/questions/answers     loop: answer
13. POST   /api/v1/presets/{id}/questions/validate
14. POST   /api/v1/presets/{id}/questions/approve     preset is now ready

15. POST   /api/v1/presets/{id}/courses               Create course under the preset
16. GET    /api/v1/courses/{id}/brief/next            loop: get brief question
17. POST   /api/v1/courses/{id}/brief/answers         loop: answer
18. POST   /api/v1/courses/{id}/brief/validate
19. POST   /api/v1/courses/{id}/launch                (Idempotency-Key required)
20. GET    /api/v1/courses/{id}/status                poll until ready
21. GET    /api/v1/courses/{id}/modules/{n}/video     fetch deliverables
Steps 1 to 14 are the one-time preset setup. Steps 15 to 21 are the per-course flow. Once a preset exists, every new course for that company is six API calls (15 to 20) plus delivery fetches.

​
Option B: No-preset flow (quick start)

For one-off courses or low-volume use, skip the preset entirely. The brand is auto-derived from the source material. 
1.  POST   /api/v1/courses/no-preset                  Create course (brand auto-derived)
2.  GET    /api/v1/courses/{id}/brief/next            loop
3.  POST   /api/v1/courses/{id}/brief/answers         loop
4.  POST   /api/v1/courses/{id}/brief/validate
5.  POST   /api/v1/courses/{id}/launch                (Idempotency-Key required)
6.  GET    /api/v1/courses/{id}/status                poll
7.  GET    /api/v1/courses/{id}/modules/{n}/video     fetch deliverables
Use this for the quickstart.

​
When to use which

Use caseRecommended path
One-off course for a single clientNo-preset
Internal proof-of-conceptNo-preset
Multiple courses for the same companyPreset
White-label LMS that serves many clientsPreset, one per client
ARESS-style scale (many companies, recurring courses)Preset, with full setup automation
You can mix and match: nothing prevents creating no-preset courses on an account that also has presets.

​
Brief vs preset questions

Two questionnaires exist in the API, do not confuse them:
  • Preset questions (Option A only) capture the company’s sales profile: industry, target buyer persona, pricing model, sales motion. Asked once per preset.
  • Brief questions (both options) capture this course’s intent: target audience, learning objectives, tone. Asked once per course.
Both use the same next + answers loop pattern.

​
Review checkpoints

By default the pipeline pauses for human approval at up to three checkpoints during generation. See Review Checkpoints for the full model.

​
Async setup operations

Several preset setup steps (brand synthesis, slide style sampling, voice style sampling) are queued. They return 202 Accepted immediately. Poll GET /api/v1/presets/{id}/setup-status until each artifact reaches ready status before proceeding. 
{
  "brand":       {"status": "ready"},
  "slide_style": {"status": "running"},
  "voice_style": {"status": "queued"}
}
Possible statuses: queued, running, ready, failed.