The challenge When you record in real-world settings, your footage often captures people who never agreed to appear — someone caught at the edge of the frame — or sensitive information shown on screen, such as personal or customer data. Today Mozaik has no way to hide these automatically, so teams have to re-shoot, avoid certain shots, or manually edit each video. For any organisation that handles sensitive information on camera, that is slow, error-prone, and a real compliance risk. Our approach Introduce automatic anonymisation. Let Mozaik detect faces and sensitive information in your footage and obscure them for you — blurring or pixelating a face, placing a sticker or overlay over a person who hasn't consented, and blending over sensitive data shown on screen. You stay in control: review what was detected and adjust what gets hidden before you finalise the video. Built for data protection This is built for privacy-sensitive and regulated environments — any GDPR-bound organisation handling personal or confidential information — so anonymisation needs to stay reliable across the whole clip, not just single frames. Expected impact Teams that handle sensitive information can produce compliant videos without manual redaction. People who didn't consent, and sensitive on-screen data, are protected automatically. Removes a compliance blocker for privacy-sensitive customers.

David Kettner 6 days ago

The challenge When you bring 16:9 (landscape) footage into a 9:16 (vertical) project, the speaking person often isn't positioned well in the automatically generated draft — they can end up off-center or partly out of frame. Today that means going in and manually re-cropping clips to get the framing right, which adds a fine-tuning step to nearly every vertical project that starts from 16:9 source material. Our approach Let's introduce a new AI agent that automatically reframes 16:9 source footage for a 9:16 project. The agent detects the speaker in the landscape frame and reframes the shot to keep them well-positioned in the vertical format — so your draft already looks right instead of one you have to correct by hand. Expected impact Less manual re-cropping — vertical drafts from 16:9 footage arrive with the speaker already well-framed A faster path from landscape source recordings to a polished 9:16 result More consistent, professional-looking framing across your clips and projects

David Kettner 6 days ago

The challenge When Mozaik's standard agent creates your lower thirds and subtitles, it often misspells the proper names in your content — company names, product names, and event names. Today the only way to teach Mozaik the correct spellings is by setting up an individual agent, so everyone working with the standard agent still has to fix these names by hand on every video. Our approach Let's let you store a glossary of your proper names together with their correct spellings — uploaded as a simple Excel or text file — and have the standard agent apply it automatically whenever it writes lower thirds and subtitles. The glossary would be available in the standard agent, not only through a custom individual agent. Expected impact Company, product, and event names are spelled correctly in lower thirds and subtitles automatically Available in the standard agent everyone uses, without building a custom agent Less manual correction and more consistent, professional-looking videos

David Kettner 6 days ago

The challenge Today only account admins can record the voiceover for a video — in many teams that's just a handful of people in the office. The protagonists who actually appear in your videos can't speak their own voiceover, even when they're the natural voice for the story. That's especially limiting when someone films something and isn't on camera themselves: there's no simple way for them to add their own narration. Our approach Let protagonists record a voiceover themselves — directly in the app or through the guest access link they already use to contribute. Recording a voiceover would become a natural part of the guest journey, right alongside the footage they capture, with no admin hand-off needed. Expected impact Protagonists and corporate influencers can voice their own clips without waiting on an admin Narration for filmed footage where the speaker isn't on camera becomes easy to capture Voiceover work is shared out to the people closest to the content instead of bottlenecking on a few admins

David Kettner 10 days ago

The challenge Right now, when a video is finished you have to download it from Mozaik and then upload or post it by hand wherever it needs to go — for example your social-media channels. That extra download-and-re-upload step costs time, is easy to forget, and breaks your flow when you're publishing videos regularly. Our approach Let's add a direct publishing integration so you can send a finished video straight from Mozaik to where you need it — your social-media platforms — without manually downloading and re-uploading it. Expected impact Publish a finished video in a few clicks instead of downloading and re-uploading it Skip a manual, error-prone step in your workflow Get your videos in front of your audience faster

David Kettner 10 days ago

The challenge Today Mozaik has no built-in avatar. If you want a presenter-style, talking-avatar video, you have to create that footage in another tool and bring it into Mozaik as raw material — Mozaik can only cut and arrange the clip you already made, not generate one for you. There's no way to type a script and have Mozaik produce a brand-new video of an avatar delivering it. Our approach Let's introduce native talking-avatar generation: pick an avatar, write or paste your script, and have Mozaik generate a brand-new video of that avatar presenting your text — ready to drop straight into your project alongside your other scenes. Expected impact Create presenter-led videos without a camera, a studio, or a separate avatar tool Turn a plain text script into a finished talking-avatar video in minutes Keep the whole workflow inside Mozaik instead of stitching in clips made elsewhere

David Kettner 10 days ago

The challenge When you build a portrait (9:16) template and then record with a webcam, your footage comes out with black bars. Webcams only capture in landscape, so the recording can't fill a portrait frame — and today nothing tells you that. You only notice the black bars after recording, without an obvious reason why. Our approach Let's add a clear heads-up — in the template and/or right before you start recording — whenever the recording format won't match the template's aspect ratio, so you know what to expect and can adjust before you record instead of after. Expected impact No more surprise black bars after a webcam recording You understand up front why webcam footage won't fill a portrait frame Fewer re-records and less back-and-forth during onboarding

David Kettner 10 days ago

The challenge When you add a text graphic to a scene (the "Graphic with text" element) the text formatting is fixed. You can't change how the headline is aligned or adjust the spacing between the letters, so there's no way to match your own layout or style. If you want a centered or right-aligned headline, or want to spread the letters out a little for a cleaner look, there's currently no option for it. It's also inconsistent with other elements: the "Video with lower third" element already lets you choose between left- and right-aligned text, which makes the missing controls in text graphics hard to understand. Our approach Let's give text graphics real text formatting controls. Add the ability to: Align the text left, centered, or right Adjust the letter spacing of the text And bring the alignment options in line across text elements so text graphics and lower thirds behave the same way. Expected impact Headlines and text match your brand and layout instead of a fixed style Center or right-align text wherever it reads better on screen Fine-tune letter spacing for a cleaner, more polished look Consistent text controls across text graphics and lower thirds

David Kettner 10 days ago

The challenge Today a fixed color and saturation setting is applied to every recording, but real footage is shot under very different lighting. The same setting looks too dark on one clip and too bright on the next, and it can leave a visible color cast such as a red tint. Fixing this by hand on every single video is slow and gives inconsistent results. Our approach Introduce an agent that analyzes each uploaded recording and automatically applies the right color and exposure correction for that specific clip — balancing brightness and neutralizing color casts like a red tint — instead of relying on one fixed setting or manual retouching per video. Expected impact Consistent color and brightness across recordings shot in different lighting No more manual color and exposure retouching on every clip Better-looking videos with less effort, even when the original footage isn't perfectly lit

David Kettner 10 days ago

The challenge When you record, the on-screen camera stencil that guides where to place yourself can only be switched on or off — its position is fixed. Presenters often find the preset frame position unnatural, which makes it harder to line themselves up and can put off less confident contributors from recording at all. Our approach Let's make the stencil position adjustable so you can move the frame to where it feels natural — for example a little lower in the picture — instead of being locked to a single preset spot. Expected impact Record in a position that feels natural instead of forcing yourself into a fixed frame Easier setup for first-time and less confident presenters More people comfortable recording, so you collect more contributions

David Kettner 10 days ago

The challenge In split-screen layouts that pair your recording with a graphic, the video sits on the left and the graphic on the right by default. Some teams want the opposite — graphic on the left, video on the right — because we read from left to right and they want viewers to land on the graphic (for example a set of bullet points) first. Right now the sides are fixed, so there's no way to switch which side the video and the graphic sit on. Our approach Let's add an option to swap the two sides of a split-screen-with-graphic layout, so you can choose whether the video or the graphic sits on the left. It's the same layout you already use, simply mirrored to match how you want people to read the scene. Expected impact Put the graphic or the video on whichever side fits your story Guide the viewer's eye in your reading direction, from left to right Keep the familiar split-screen layout — just mirrored, with no need to rebuild the scene

David Kettner 10 days ago

The challenge Today the font in your videos is fixed by the design or brand kit you're using — it's baked in, so you can't change the typeface on your own. If you want a different font, you currently have to go back through our team to have it swapped in the brand kit. That makes even a simple typeface change slow and dependent on us. Our approach Let's let you change the font in your brand kit yourself. You'd be able to pick the typeface directly in your brand kit settings and have it apply across your videos — no round-trip through our team. Expected impact Change your brand font on your own without waiting on our team Keep your videos on-brand as your typography evolves Update the font in one place and have it apply consistently

David Kettner 10 days ago

The challenge Right now, when you add background music to a project it applies to the whole video and plays across every clip from start to finish. There's no way to give one part of the video its own music. For example, you might want a specific melody under just your intro — the clip with your logo reveal — and something different, or nothing at all, for the rest. Today the single track you pick runs over everything. Our approach Let's make background music work at the clip and scene level, not only for the whole video. You'd be able to assign a specific track to an individual clip or scene — for example a short melody under your intro and logo reveal — while the rest of the video keeps its own music or stays silent, with smooth transitions between the different pieces. Expected impact Give your intro, logo reveal, or any single clip its own music Combine several tracks in one video instead of a single song throughout More control to match the music to the mood of each part of your video

David Kettner 10 days ago

The challenge In the editor's Effects tab you can adjust basic color correction today, but there's no way to blur a clip or a still image inside Mozaik. A common need is to soften the freeze-frame at the start of a clip so you can place a question or title on top of it and keep all the focus on the text — but without a blur effect the background stays sharp and competes with the overlay. A zoom effect is on the way, but blur isn't available yet, so the only workaround is to blur footage in an external tool and re-import it. Our approach Add a blur effect in the Effects tab that you can apply to any clip — both video clips and image or still-image clips — right alongside the existing color-correction adjustments. Let you control the strength of the blur, from a soft, subtle blur to a heavy one, for example to sit a question or title cleanly over a blurred freeze-frame. Expected impact Blur any clip or still image directly in Mozaik, without external tools Make text and question overlays stand out by softening the background behind them Round out the Effects tab alongside color correction and the upcoming zoom effect

David Kettner 10 days ago

The challenge When you duplicate a template, the automation workflow connected to it doesn’t come along — the workflow has to be copied over separately. This trips people up most during onboarding: you’re encouraged to duplicate the prepared template and experiment in your own copy, but because the automation didn’t carry over, that first run produces no automatic result and it looks like something is broken. Our approach Let’s carry the connected automation workflow along when you duplicate a template, so the copy is ready to produce automated results right away — just like the original — without having to rebuild or copy the workflow by hand. Expected impact Duplicated templates work end to end immediately, including their automation A smoother first experience in onboarding, with no “is it broken?” moment No need to copy the automation workflow over separately

David Kettner 10 days ago

The challenge Today your branding — like your logo and brand colours — applies to a whole project, so every scene in a video shares the same settings. But a single video often needs to flex: you might want one logo in one scene and a different logo in another, or adjust the text colour in a scene where the background is especially light or dark so your text stays readable. Right now there's no way to vary these branding settings from one scene to the next within the same project. Our approach Let's let you define branding at the scene level. Within a project you'd be able to pick a different logo for individual scenes and adjust scene-specific settings such as text colours, so each scene can look exactly the way it needs to while staying within your brand. Expected impact Use different logos across the scenes of a single video Adjust text colours per scene so text stays readable on light or dark backgrounds Keep every scene on-brand without splitting a video into separate projects

David Kettner 10 days ago

The challenge When projects are created automatically — for example one per recipient from a template — their names are built from the template title followed by the person's name. When template titles are long, every project in your overview starts with the same long title, so you can't tell at a glance whose project is whose and end up opening each one just to find the name. Our approach Let's make the order of the automatically generated project name configurable, so you can choose to show the recipient's name first, followed by the template title — or keep the current order. Putting the most distinguishing part, the person's name, up front keeps your project list scannable even with long template titles. Expected impact See whose project is whose at a glance in the project overview No need to open individual projects just to read the name Cleaner, more scannable project lists when template titles are long Flexibility to order names to match how your team works

David Kettner 10 days ago

The challenge Right now, moving from one scene to the next is limited to your brand kit's built-in transitions or a hard cut. If you want a softer or more dynamic scene change — like a wipe — there's no way to control how that change looks. For many videos this makes the flow feel abrupt and leaves little room for a polished, personal style. Our approach We want to introduce a set of real scene transitions you can choose from and apply between your scenes, such as wipes and similar effects, alongside the existing branded transitions and hard cut. You'd pick the transition that fits each scene change and preview it before rendering. Expected impact More control over how your videos flow from scene to scene. Smoother, more professional-looking scene changes beyond hard cuts. Creative variety to match your video's style and brand.

David Kettner 10 days ago

The challenge Right now the asset pool only accepts images and videos, so there is no way to leave written guidance inside a folder. The only tool you have to point people to the right place is the folder name itself — which is why folders end up named things like "Upload raw material for AI here". When a folder is connected to an automation, everyone using it has to remember exactly what belongs there and how it should be used, because nothing inside the folder explains it. That is easy to get wrong and hard to scale as you add more folders and more people. Our approach Let's add support for text files — for example a "Readme" — in the asset pool, so you can drop a short set of instructions directly into a folder. Anyone who opens that folder can read what it is for and how to use it, right where they need it, instead of relying on the folder name alone. Expected impact Explain how a folder should be used, right inside the folder Give clear guidance for folders connected to an automation, so the right material lands in the right place Fewer mistakes and less back-and-forth from people guessing what belongs where Onboard guests and colleagues faster, without them having to memorize your setup

David Kettner 10 days ago

The challenge When you record videos — for example capturing your screen and webcam together — you often want to place them in front of a background image in your brand colours. Right now you have to build and maintain that background yourself, and it easily drifts from the rest of your Mozaik branding, so your recordings don't quite match your finished videos. A background image isn't a standard part of your brand kit today, so there's nothing ready-made that lines up with your other brand elements. Our approach Let's make a branded background image a regular element of your brand kit — matching your brand colours and the rest of your kit. You'd have a ready-to-use background that fits screen-and-webcam recordings out of the box, including when you record directly inside Mozaik, so everything you record stays consistent with what you create in Mozaik. Expected impact A background that matches your brand kit, so your recordings look consistent with your Mozaik videos No need to build and maintain your own background image Especially useful for screen-recording formats (screen plus webcam), which come up across many teams

David Kettner 10 days ago