Case 01

Automating a manual video-editing workflow for educational content at scale

A slow, editor-by-editor Premiere Pro process turned into an automated pipeline that cleans, reframes, and AI-edits recorded lesson videos in minutes — built to run hundreds of videos at volume.

Role: Managed service·Timeline: Ongoing production·Status: 25+ videos processed·blackcamel.productions →
01
Ingestraw lesson recordings in bulk
02
Visual cleanuptoolbars, masks, re-framing
03
Adaptive reframecontent-aware, no collisions
04
AI content cleanupsilence / filler / repeats
05
Render + deliverbrand-consistent finished video

The challenge

Black Camel Productions produces a high volume of recorded lesson videos for an educational content client — screen-and-webcam recordings of a teacher walking through slides. Every video went through the same repetitive manual edit in Premiere Pro:

  • Cropping recording toolbars and UI clutter
  • Scaling and re-framing the screen content
  • Enlarging and repositioning the presenter's webcam bubble
  • Masking grey page-breaks and margins
  • Exporting to a fixed broadcast spec

On top of that, an editor had to listen through each video to remove silences, filler words, coughs, and repeated or stumbled lines — the most time-consuming part of all.

The problem wasn't quality — it was throughput. Each video cost an editor roughly an hour of hands-on time, the process didn't scale, and quality drifted between editors. At volume, the manual content-cleanup pass became the bottleneck.

The solution

We built an automated video-production pipeline that reproduces the manual editing standard programmatically — plus an AI-assisted content-cleanup layer the manual process couldn't do efficiently.

It runs as a managed service: Black Camel sends the raw recordings, the pipeline processes them, and a finished, brand-consistent video comes back — ready for their team to drop into final assembly.

1. Pixel-accurate brand reproduction

We reverse-engineered the client's exact editing standard from their original Premiere presets and reference videos, so output matches their spec to the pixel — webcam framing, spacing, and scale measured and locked (for example, a 42px webcam margin matched against a reference video, not guessed).

2. Automated visual cleanup

Toolbar and UI removal, page-break masking, screen re-framing, and presenter-bubble enlargement happen automatically — no manual masking or keyframing.

3. Adaptive, content-aware framing

The system inspects each slide and adjusts automatically so on-screen images and tables are never cut off and never collide with the presenter overlay — solving real client feedback without per-video manual work.

4. AI-powered content cleanup

  • Silence and dead-air removal using voice-activity detection
  • Filler-word, cough, and breath cleanup via speech AI, with safeguards that protect actual speech from being clipped
  • Repeated-line and stumble detection using AI transcription plus a text-analysis layer — it finds where the presenter repeated themselves and proposes a tight, word-level cut, leaving a human to approve

Technical highlights

Reference-driven accuracy

Brand geometry was decoded from the client's own design assets and measured against a reference video, not estimated — so output is consistent and on-spec every time, eliminating editor-to-editor drift.

Deterministic repeat detection

For finding repeated lines, the pipeline pairs Whisper transcription (with word-level timestamps) with a text-matching layer tuned for the task — locating duplicated speech precisely and proposing a clean cut, with a human approving the final call.

Optimised rendering

We tuned the render engine for Apple Silicon with smart frame-reuse, roughly halving render time — from about 6 minutes down to ~3 minutes for a typical 15-minute lesson.

Resilient at volume

Batch processing with automatic retries on network hiccups, persisted results, and visible warnings so nothing fails silently across a large batch.

Results & impact

~3 min
Automated processing per video (was ~1 hr manual)
25+
Videos processed across batches
~2×
Faster rendering after Apple Silicon tuning
100s
Built to run at volume
  • From ~1 hour of manual editing per video to ~3 minutes of automated processing — freeing editors from repetitive work
  • Brand-consistent output — webcam, layout, and framing matched to the client's reference standard, eliminating editor-to-editor drift
  • The hardest manual task — listening for silences, fillers, and repeated lines — is now an automated first pass with human approval, instead of a from-scratch manual job

The outcome

Black Camel moved from a labour-bound, one-editor-per-video process to a scalable production line: send the recordings in bulk, let the pipeline do the heavy, repetitive work, and apply human judgement only where it adds value.

The result is lower cost per video, faster turnaround, and consistent brand quality at volume — turning video production from a bottleneck into a throughput advantage.

Delivered for Black Camel Productions; the educational end-client is kept anonymous by agreement.

Tech stack

LanguagePython
Video processingFFmpeg · OpenCV
API / OrchestrationFastAPI
TranscriptionWhisper (word-level timestamps)
Speech cleanupCleanvoice
LLM / Text analysisGemini
OptimisationApple Silicon render tuning

Similar challenge?

Tell us what you're building — we'd be glad to talk it through.

Let's talk