Avatar Video Generation

SpiritMe produces a video of a digital avatar speaking supplied text. When a script and recipient list arrive, a Beam agent submits each render within the approved avatar and template set, reads the finished file, and attaches it to the matching campaign record. Completed videos ready inside the rules are queued for sending. Renders using an unapproved avatar, script, or brand element are held and handed to a marketer, so nothing off-brand reaches a recipient. Failed or corrupted renders are also flagged for a person rather than sent broken to a contact.

Personalized Video Rendering

SpiritMe swaps names and details to tailor a base video per recipient. On a personalization run, the agent reads the recipient data, applies an approved rule on which fields to merge, and produces one video each. Videos that pass a field-completeness check are linked to the contact and the send is scheduled. Records with missing or odd merge fields, like a blank name or a placeholder that never filled, are routed to a human before anything goes out, so a recipient never opens a clip addressed to nobody or greeting the wrong person.

Render Status Notifications

Video jobs take time and can fail. A Beam agent watches the queue, reads each job's status, and updates the linked campaign task as renders finish. When a batch completes within the agreed window, the owner is notified that assets are ready. Jobs that stall past the expected time, error out, or return a file that fails a basic quality rule are escalated to a producer instead of being marked done, so a campaign owner is never told videos are ready when some quietly failed to render.