Owner 2026-09-06: we also need to update our videos; each product should get a five minute video that we publish.
Standard for every cut (from edgentx.github.io#1 and the recorded lessons): five minutes, narration.yaml is the single source for spoken words and captions (Kokoro af_heart on the .148 rig, edgent-lms tools/training-videos pipeline), visible emphasized cursor and region highlight synced to narration, shot in short passes on an open demonstration tenant with synthetic data, frames extracted and looked at before muxing, name the thing not its address, burned captions plus a VTT track (Section 508), published on edgentx.github.io with a poster and a stable URL, then linked from the product landing page and the capability statement.
One child issue per product, each owned in its product repo:
- VForce360 (autonomous AI workforce): re-cut of the existing walkthrough to five minutes with the pointer and highlight standard
- VForce Flow (edgentx/vforce-flow): build a flow node by node, run it, show the run output and the event trail
- VForce Lakehouse and Mnemo (edgentx/vforce-lakehouse): catalog, one table in full, natural language analyst with SQL shown, provenance
- Askura (edgentx/askura): build a survey, respond on a phone, watch the analysis and the report
- Edgent Learn LMS (edgentx/edgent-lms): author a course, enroll, learn on a phone, the record in the lakehouse
- Cockpit (edgentx/cockpit): an opportunity from intake through analysis, funnel score, response, and the record
Definition of done for the epic: six videos live on edgentx.github.io, each under five minutes, each linked from its product page, the canonical public demo link list updated in vforce-memory.
Owner 2026-09-06: we also need to update our videos; each product should get a five minute video that we publish.
Standard for every cut (from edgentx.github.io#1 and the recorded lessons): five minutes, narration.yaml is the single source for spoken words and captions (Kokoro af_heart on the .148 rig, edgent-lms tools/training-videos pipeline), visible emphasized cursor and region highlight synced to narration, shot in short passes on an open demonstration tenant with synthetic data, frames extracted and looked at before muxing, name the thing not its address, burned captions plus a VTT track (Section 508), published on edgentx.github.io with a poster and a stable URL, then linked from the product landing page and the capability statement.
One child issue per product, each owned in its product repo:
Definition of done for the epic: six videos live on edgentx.github.io, each under five minutes, each linked from its product page, the canonical public demo link list updated in vforce-memory.