Solution · Video and podcast
AI video for business: one brief, three formats, four languages
An AI agent writes the script, breaks it into a storyboard, edits the video with your visuals, then adds the voice-over and subtitles. It also turns your recordings into podcast episodes. A person approves everything before it goes out.
- A synthetic voice, or yours with your consent
- Subtitles in FR · DE · EN · IT
- Nothing published without approval
What an AI video and podcast agent does
Video is everywhere: LinkedIn, Instagram, YouTube, shop-window screens. For a small business, the problem isn’t ideas, it’s time: script, images, voice, editing, subtitles, then all of it again in German or in vertical format. A 30-second video can keep several people busy for days.
An AI video agent takes over the production work. From a brief, it writes the script, proposes a scene-by-scene storyboard and assembles a motion-design edit with your logo, brand colours and photos, topped up with licensed footage and music. It adds an AI voice-over (a synthetic voice, or your own if you consent) and subtitles in French, German, English or Italian, then delivers the video in 9:16, 1:1 or 16:9.
For AI podcast production, the pipeline starts with your recording: transcription with Whisper, clean-up, chapters, show notes, short clips and social posts. You keep your voice and your guests; the agent handles everything that comes after the microphone.
Nothing is published without your say-so. Someone on your team reviews every video and every episode, and we label AI-generated content where relevant. We don’t promise photorealistic avatars: we make clear videos in your brand colours that say the right thing.
The problem
What holds video back today
Too slow to produce
Each video needs different skills and several rounds of feedback. So it slips behind everything else.
One language at a time
In Switzerland, a video in French only reaches part of your audience. Redoing the voice and subtitles in German or Italian doubles the work.
The podcast stops at the microphone
The recording is done, but the show notes, chapters and clips never follow. The episode stays hard to find.
Interactive demo
Try the AI video studio
Pick a fictional brief, a format and the voice language. The agent writes, storyboards, voices, subtitles and renders the video: watch it in the player, then approve it. In the Podcast tab, a raw recording becomes a complete episode.
AI video and podcast studio
Writes, storyboards, voices, subtitles and renders. You approve before anything goes live.
Pick a brief, a format and the voice language, then run the agent. Everything is fictional and nothing is published.
Engine
The brief
Régie Pervalis SA (fictional)
- Goal
- Announce the new tenant app and get people to download it.
- Key message
- Report a problem in 30 seconds, photo included.
- Audience
- The property manager’s tenants
- Your files
- Logo, brand colours, 3 app screenshots
- Call to action
- Download the app
Script
AI queuedRun the agent: the script is written here, scene by scene.
Storyboard
AI queuedThe 5 storyboard scenes will appear here.
Preview
Tool queuedThe preview comes to life during production.
CSS demo · no sound · fictional dataSubtitles
Code queued- After the voice-over, code cuts and times the subtitles.
Timeline
AI queuedAfter the render, click in the timeline to move around.
Approval and publishing
Human queuedAfter the render, the video waits here for your approval.
Everything runs in your browser: nothing is sent or stored. Simplified demo: answers are prepared in advance or computed by simple rules in your browser. In production, a language model suited to your data does this work, and someone on your team signs off. Here the video is drawn in CSS, with no sound and no real file. In production: your visuals, a real voice-over and an MP4 file.
I want this for my businessWhat’s happening, step by step
-
The AI writes the script AI
From your brief, a language model writes a short text, split into scenes, in your brand’s tone of voice.
-
It builds the storyboard AI
Each scene gets a visual (your logo, your photos, licensed footage) and a duration.
-
The voice-over is generated AI
A synthetic voice reads the script in the language you chose. Your own voice is possible, only with your consent.
-
Code times the subtitles Code
No AI here: code splits the text, syncs it with the voice and checks readability (line length, reading speed).
-
The video is rendered Tool
An editing tool puts pictures, voice, music and subtitles together into one video file, in the format you need.
-
You approve Human
You watch, ask for a change or approve. Nothing is published without your say-so.
Run the demo: each step lights up here.
ModelApertus 70B, Whisper and open-source voices
Data processedA server in Switzerland or yours
Picture a full production team (scriptwriter, editor, voice-over artist, subtitler) that turns a three-line brief into a video ready for review, in your languages and formats. Then it waits for your go-ahead.
Words to know
- Storyboard
- A video broken down into scenes: what you see, what you hear, and for how long.
- Synthetic voice
- A voice generated by software from text (text-to-speech). It can read the same script in several languages.
- SRT file
- A plain text file with the lines and their timings, which platforms such as YouTube accept. The text can also be burned into the picture.
- Whisper
- An open-source speech recognition model: it turns a recording into timestamped text, in many languages.
Why it’s safe
- Nothing is published without your OK: a person approves the video before release.
- Subtitles and durations are checked by code.
- A real person’s voice is used only with their written consent.
- You choose where the AI and your data run.
The agent’s pipeline
- Input Brief and your files Brief text, logo, brand colours, photos, screenshots. For a podcast: the raw recording.
- AI Script and storyboard The model returns schema-validated JSON: scenes, on-screen text, voice-over, visual, target duration.
- AI Voice-over Text-to-speech scene by scene; the real length of the audio sets the edit.
- Code Subtitles and checks Cut into SRT and WebVTT, 2 lines at most, reading speed and overlaps checked.
- Tool Render FFmpeg puts together animated templates, licensed footage, voice, music and subtitles; loudness normalised (−14 LUFS).
- Human Approval Preview link and comments per scene. Publishing is only scheduled afterwards.
AI reads and writes. Code checks. Humans decide.
Stack
Guardrails
- Rights checked: footage and music come from libraries with a commercial licence or from your own files, and every source is recorded.
- Voices with consent: a real person’s voice is only used with their explicit consent, which they can withdraw.
- Transparency: an “AI-generated voice” or “AI-generated content” notice where relevant or required, for example by a platform’s rules or, for EU audiences, by the AI Act. See Swiss AI law.
- Checks by code: subtitles are checked (line length, reading speed, overlaps) before the render.
- Nothing goes out on its own: publishing is only scheduled after a human approves.
What the agent produces (JSON)
{
"video_id": "pervalis-launch-en",
"format": "9:16",
"language": "en",
"duration_s": 30.1,
"scenes": [
{
"n": 1,
"type": "kinetic_text",
"on_screen": "Trouble at home?",
"voice": "A dripping tap, a stuck blind? …",
"duration_s": 7.2,
"source": "motion_design"
},
{
"n": 2,
"type": "app_mockup",
"on_screen": "Report it in 30 seconds",
"duration_s": 6.4,
"source": "client_screenshots"
},
{
"n": 4,
"type": "cutaway",
"duration_s": 4.9,
"source": "licensed_stock",
"licence_ref": "STK-48213"
}
],
"voice": {
"type": "synthetic",
"real_person": false,
"disclosure": "end_card"
},
"subtitles": {
"files": [
"fr.srt",
"de.srt",
"en.srt",
"it.srt"
],
"max_chars_per_line": 32,
"max_cps": 15.3
},
"status": "awaiting_approval"
}Choose the engine
How we set it up
- Day 1 Free audit 30 minutes to look at your content, channels, languages and what must stay human.
- Weeks 1–2 Pilot A first video and a first episode from your own files, to set the tone, the templates and the voice.
- Weeks 2–4 Go-live Templates in your brand colours, the chosen voice, approval rules and a publishing plan.
- Every month Production and care New content, adjustments and a monthly report.
What we need from you
- Your logo, brand colours and a few photos or videos
- A brief per piece, or your recordings for the podcast
- To use your voice: a reference recording and your written consent
- The person who approves before publishing
What we set up
A full production line, in your brand colours and your languages
Templates in your brand, your approval rules, your channels. You stay in control of every post.
Script and storyboard
From a brief or an article, a short script broken into scenes: on-screen text, voice-over, visual, duration.
Motion-design editing
Animated templates in your brand colours, with your logo and photos, plus licensed footage and music.
AI voice-over
A natural synthetic voice, or your own voice, only with the explicit consent of the person.
Subtitles in four languages
Burned into the picture or as SRT and VTT files, in French, German, English and Italian, checked by code.
Podcast production
Transcription with Whisper, clean-up, chapters, show notes, clips and social posts.
Formats and publishing
Every piece in 9:16, 1:1 and 16:9, with a publishing plan scheduled in your tools after you approve.
How we work
How it goes
-
Free audit
In 30 minutes we look at your content, your channels, your languages and what must stay in human hands.
-
Pilot on one piece
A first video and a first episode from your own files, to set the tone, the templates and the voice.
-
Go-live
Templates, voice, approval rules and publishing plan are in place. We aim for 2 to 4 weeks after the audit.
-
Production and care
We monitor quality, adjust the templates and send you a report every month.
Our commitments
Content you can put your name to
A human approves
By default, no video or episode is published without approval. You can ask for changes scene by scene.
Rights and voices under control
Footage and music are licensed or supplied by you. A real person’s voice is only used with their consent.
Transparency about AI
An “AI-generated” notice where relevant or required (see our guide Swiss AI law). For internal content, Swiss hosting is available.
Frequently asked questions
Can AI make a video from plain text?
Yes: from a brief or an article, the agent writes the script, proposes a storyboard and edits a motion-design video with your visuals. A person approves each video before it’s published.
Do you offer photorealistic AI avatars?
No. We make motion-design videos from your visuals, licensed footage and music, with a synthetic voice or your own. There is no doubt about what is real.
Can you use my voice, or a colleague’s?
Yes, but only with the explicit consent of the person concerned. A voice is personal data: we agree in writing how it will be used, and the person can withdraw their consent. Otherwise, we use a synthetic voice.
Do I have to say a video was made with AI?
Sometimes, and we do it where relevant. Switzerland has no AI-specific law yet, but the EU AI Act requires deepfakes shown in the EU to be disclosed, and several platforms ask creators to label realistic AI-generated content. See our guide Swiss AI law.
How does AI podcast production work?
You record, the agent does the rest: transcription with Whisper, clean-up of the text and sound, timestamped chapters, show notes, short clips and posts. A person reviews everything before it’s published.
Which languages and formats?
French, German, English and Italian, in 9:16, 1:1 and 16:9. Subtitles come burned into the picture or as SRT and VTT files. For each language, we recommend a review by someone who speaks it.
How much does it cost?
The price depends on the volume of content, the languages and the tools to connect. The audit is free; you then get a clear quote with a set-up cost and a monthly fee. Request a quote.
NotHuman
How many videos are still stuck in your head?
In a free 30-minute audit, we look at your content and channels and suggest a first format to produce.