The context
A lesson recording is not yet ready-to-use learning material. It needs to be stored, assigned to the right course, published and made accessible. One lesson may have several video files; transcripts and summaries can arrive at different times.
The work for AIV treats these steps as one process. My contribution covers the design, integration and implementation of the workflow, keeping the outcome of each operation visible.
The problem
An uploaded video does not prove that a lesson is available in the right place. A successful publication does not confirm that the backup, transcript or every part of the recording has been handled.
The process needs a consistent relationship between recording events, the learning destination and the status of each step.
Design decisions
The first decision is to separate a lesson’s identity from its individual files. One recording can have several parts that must be ordered and published together. Two lessons on the same day must not be combined simply because they share a date.
The second is to treat videos, transcripts and summaries as materials that may become available at different times. The process must associate them with the same lesson without relying on a fixed arrival order.
The third is to keep outcomes separate: archiving, video upload and publication have their own checks. The final notification needs to make that distinction clear.
The implemented system
The workflow connects Zoom recordings, Google Drive archiving, Bunny video distribution and FluentCommunity publication. Operational notifications in Discord help people read the outcome and identify steps that need attention.
Recordings with multiple files are ordered and published as parts of one lesson. Transcripts and summaries are associated with the corresponding recording. Manual routing supports cases that need an operator’s decision.
AI has a specific role in content tasks. Video enrichment is requested when the required material is available and the video has finished processing.
Results and limits
The documented result is a process connecting archiving, publication and operational checks, including multi-file lessons and materials that arrive later.
This case study does not claim a percentage saving or a number of hours recovered. Measuring those outcomes requires a manual-work baseline, an observation period and the cost of supervision. A video’s duration or a workflow’s running time is not equivalent to a person’s time.
Beyond delivery
Exceptions remain part of the system: incomplete recordings, late materials, destinations awaiting assignment and unavailable external services. Improvement also means making these conditions visible and manageable.
This is where the cycle starts again: observe the real work, choose a useful measure and decide what to improve next.