Commercial Video Editor & Post-Production Lead; AI Evaluation
-
Creative Arts/Media
Job Description & How to Apply Below
$70–120/hr
Mercor
Full time, 40 hours a week
Lead commercial video editing and post‑production for AI model training and evaluation.
What you would do- Design and produce editorial tasks that mimic agency workflows
- Edit reference cuts in Premiere Pro or DaVinci Resolve
- Assess AI‑generated video outputs and document issues
- Create guidelines, rubrics, and technical checks for editing quality
- Coordinate with research and SME teams to align training data
- 5+ years delivering commercial or agency edits with brand credits
- Daily expert use of Premiere Pro and DaVinci Resolve; familiarity with Avid and After Effects
- Proficiency with source file formats, Pro Res/DNxHR, FFmpeg, and color grading pipelines
- Experience cutting broadcast, online, and social versions across multiple aspect ratios
- Ability to work full‑time (40h/week) and communicate decisions clearly in writing
AI edit critique
“A client provides a 30‑second AI‑generated hero cut that drifts off‑beat at the 12‑second mark; how would you document the issue and suggest a fix?”
Why this comes up:
They need to see you can spot subtle pacing or sync errors that affect model learning, showing both creative and technical acuity.Task design for agency workflows
“You must create a set of cutdowns (60s to 6s) and 9:16 reframes for a new sneaker campaign; what steps do you take to structure the project files and deliverables?”
Why this comes up:
Designing realistic editing challenges ensures the AI sees diverse, production‑ready scenarios, which is critical for robust training data.File format and transcode handling
“Given a batch of .prproj files with mixed codecs, how would you use FFmpeg to generate consistent Pro Res proxies while preserving timeline metadata?”
Why this comes up:
Accurate handling of source and delivery formats prevents data corruption and maintains visual fidelity for AI training.Color grading consistency
“A brand requires a specific LUT for skin tone consistency across all social cuts; describe how you apply and verify the LUT in Resolve for multiple versions.”
Why this comes up:
Consistent Rec.
709 grading teaches the model proper color standards, especially for skin tones and brand palettes.Collaboration with research teams
“The research team reports AI struggles with subtitle timing on a localized version; how would you adjust your editing workflow to address this?”
Why this comes up:
Aligning editorial decisions with research goals ensures that feedback directly improves model performance.Guideline and rubric creation
“You need to draft a rubric for evaluating AI‑generated title safety; what criteria would you include and how would you weight them?”
Why this comes up:
Clear standards let multiple reviewers evaluate AI outputs uniformly, accelerating model iteration.
Candidates may be given a raw footage set and asked to produce a 30‑second hero cut, a 6‑second cutdown, and a 9:16 reframed version, then write a brief critique of an AI‑generated version of the same material.
How to prepare- Assemble a portfolio that includes agency cutdowns and aspect‑ratio reframes with detailed project files
- Practice writing concise, technical feedback on AI‑generated edits, focusing on pacing, sync, and color
- Refresh command‑line FFmpeg commands for batch transcodes and metadata extraction
- Study common AI video tools (Runway, Luma, etc.) and note their typical failure modes
Pay $70–120/hr
Hours Full time, 40 hours a week
Where Remote, United States
Open to USA
(If this job is in fact in your jurisdiction, then you may be using a Proxy or VPN to access this site, and to progress further, you should change your connectivity to another mobile device or PC).
View / Apply for Jobs
Matching My Jurisdiction