EditDNA
"Turn Raw Content Into Edited Video Drafts."
Build an AI editing assistant that understands a script, raw video, reference style, and asset folder, then creates an edit plan, b-roll mapping, captions, typography moments, timeline JSON, and optionally a rendered short-form draft.
The Problem Statement
Creators, founders, agencies, and businesses often know what kind of video they want, but editing requires many micro-decisions: where to cut, where to add b-roll, which words to highlight, what captions to use, when to zoom, and how to match a reference style. The goal is to automate editing intelligence, not just generate videos.
What You Need To Build
Start with the core requirements, then layer advanced features and bonus innovations to stand out.
Must-Have
Minimum features required for a valid submission.
- Script input
- Raw/AI avatar video upload
- Asset folder upload
- Basic reference style or prompt input
- Script-to-scene breakdown
- B-roll recommendation for each script beat
- Caption and typography plan
- Timeline JSON generation
- Basic preview or render plan
Should-Have
Features that make the prototype stronger.
- Reference video analyzer
- Auto caption timing
- B-roll matching from uploaded assets
- Typography styling
- Motion graphic suggestions
- Basic Remotion/FFmpeg rendered output
- Multiple edit versions
Bonus
High-impact additions for standout teams.
- Long video to 3 short clips
- Hook detection
- Dead-space removal
- AI-generated b-roll prompts
- Voice-based edit commands
- Brand style memory
- Export-ready short-form video
Suggested User Flow
Input
User uploads a 1-minute raw talking-head video and a folder of 10 b-roll clips.
Analysis
AI transcribes video, detects pauses, and maps sentences to b-roll context.
Edit Plan Generation
System generates a timeline JSON with cuts, b-roll placements, and captions.
Render/Preview
User previews the rough draft in a web-based player.
Expected Deliverables
- Working prototype
- GitHub repository
- README with setup instructions
- Demo video (max 3 minutes)
- Architecture diagram
- Dataset/API explanation
- Known limitations
Judging Criteria
Suggested Tech Stack
* These are suggestions only. You are free to use any modern stack you prefer.
Rules & Compliance
Participants must use copyright-safe sample assets, self-created assets, or assets provided for the hackathon. Do not copy another creator’s exact content, watermark, brand identity, or copyrighted footage. The system may learn editing patterns, but must not clone protected creative work directly.
- Use original work
- Do not copy copyrighted assets
- Do not use unauthorized scraping
- Do not collect private data
- Mention all APIs/data sources
- Open-source allowed with credit
- AI tools allowed, but explain usage
EditDNA track FAQ
Direct answers about the challenge scope and expected prototype.
What is EditDNA?
EditDNA is an XE Labs Build Sprint 2026 challenge track for an AI video editing automation engine. It focuses on editing intelligence: understanding a script and raw footage, recommending cuts and b-roll, planning captions and typography, generating a structured timeline, and optionally producing a rough short-form video draft.
What should teams build in EditDNA?
Teams should build a working prototype that accepts a script, raw or avatar video, supporting assets, and a reference style or prompt. It should break the content into scenes, recommend b-roll and captions, and produce timeline JSON or an equivalent edit plan. A usable preview or basic render makes the submission stronger.
Can EditDNA copy a creator’s editing style exactly?
No. A prototype may identify general editing patterns, pacing, caption treatments, or shot types, but it should not clone another creator’s protected work, watermark, brand identity, or copyrighted footage. Teams should use self-created, licensed, provided, or copyright-safe assets and document the source and limitations of reference material.
Ready to build this track?
Strong submissions and finalists may be invited for paid internship interviews to continue production development under XE Labs.