Use case · Webinar repurposing
One webinar.
A campaign kit.
The webinar took six weeks to fill and fifty-eight minutes to run - then it became a link in a follow-up email. Webinar to short clips AI, done as composition, treats the recording as campaign material: brief the promo cut, the speaker moments and the argument summary from the same session, and render each one captioned and on brand.
Every deliverable is grounded in what was actually said, with source timecodes to show it.
One video or an entire archive.
Webinar repurposing product demonstration
The problem
Fifty-eight minutes, one link.
B2B webinars are the most expensive recordings most marketing teams make - speakers booked, questions gathered, an hour of genuine argument - and the standard afterlife is a gated replay that a fraction of registrants ever open. The session said plenty; the campaign shipped almost none of it.
Repurposing stalls on production and dies on consistency: three clips cut by three people over three weeks drift in caption style, framing and sign-off. By the time the assets exist, the moment the webinar spoke to has moved on.
Campaign workflows are the daily work of marketing teams - their page is here →
The kit
Three briefs, one session.
A campaign kit is not one edit repeated at three lengths. Each piece has its own job, so each gets its own brief against the same indexed session - these three, verbatim.
The promo cut
“Thirty seconds to make a cold viewer register for the replay: open on Dana's sharpest claim about buying committees, cut to the one number that backs it, end mid-argument.”
target_duration_s 30 · 9:16 · crop_mode "stack"
The speaker moments
“Forty-five seconds of Priya only: her three strongest moments on procurement, each complete enough to stand alone.”
target_duration_s 45 · 1:1
The argument summary
“The session's argument in two and a half minutes, in the order it was made: problem, evidence, objection, answer. Both voices.”
target_duration_s 150 · 16:9 · chronology "strict"
Each brief is one request - POST /v1/cuts with the session, a target length and the formats - and each plan is reviewed before it renders. The full request surface is documented with the API; what composition itself means is defined on the product page.
Brand kits
On brand, by configuration.
A brand kit is set up once: your caption typeface and colours, a logo, an end card per aspect ratio. Name the kit on a cut and every render applies it - captions burned in and styled to your type at your chosen size and margin, the logo watermarked top-right, the end card appended as a two-second still.
The surface ends there, stated exactly: intro cards, lower thirds, music and colour grades stay out of scope. What a kit guarantees is that captions, mark and sign-off match across every clip the webinar produces, whoever briefs them and whenever.
Speakers get the same once-only treatment. Diarisation separates the session's voices and you attach names a single time - which is why “Priya only” is an instruction the platform can execute, with her talk time visible before you brief.
Where it fits
The conditions, printed.
CHOOSE IT WHEN
- Your webinars are argument-led and speaker-led - the value is in what gets said.
- A campaign needs several formats fast: each brief renders up to three aspect ratios from one plan.
- Consistency across clips matters more than per-clip artistry - same captions, same mark, same sign-off, whoever briefs them.
- Each clip must trace back to the session: source timecodes on every segment, an EDL with every render.
LOOK ELSEWHERE WHEN
- The value sits on the slides rather than in the speech - retrieval runs on the transcript, and a vertical crop cannot rescue a dense chart.
- You want animated intros, music beds or motion graphics; those sit outside what a brand kit applies.
- The clip must follow the active speaker around the frame. That crop is on the roadmap; vertical today is bars or stacked crops.
THE EDGES, PLAINLY
A source file caps at 16 GB and four hours - comfortably past any webinar, tight for an all-day summit, which loads as separate sessions. Planning is asynchronous: each brief is accepted at once and its plan follows when composed, so a three-piece kit is three plans you read, then render. Source minutes are counted once, as you bring footage in; briefing, plans, revisions and renders use none. Plans and prices are on pricing.
The objection
A fair challenge, met.
- Our webinar platform already auto-generates clips - why write briefs?
- Auto-clips answer “what stood out”; a campaign needs “what convinces this audience at this stage”. A brief carries that intent - the claim, the audience, the ending - and the plan shows exactly which sentences will carry it, with source timecodes, before anything renders. When the first plan misses, you say so in plain language and read the next version. What ships is the argument you chose, in your brand's dress, on every channel at once.
Questions
- How do I turn a webinar into short clips with AI?
- Upload the recording as a session; it is transcribed with word-level timecodes and its speakers are separated. Then write one brief per clip - what it must say, who it is for, a target length from 10 to 600 seconds, up to three aspect ratios. Review each plan against its source timecodes, then render; every clip ships with SRT captions, an EDL in JSON and a QC report.
- Can AI cut clips of one speaker from a webinar?
- Yes. Each session's voices are diarised and you attach display names once, so a brief can ask for a single speaker's strongest moments - and the plan shows which segments belong to whom, with talk time per speaker visible before you even brief.
- How do I keep webinar clips on brand?
- With a brand kit: caption typeface and colours, a logo and per-aspect end cards, configured once and named on the cut. Every render burns in captions styled to the kit, watermarks the logo top-right and appends the end card as a two-second still - the same treatment on every clip, from every brief, by anyone on the team.