# Why We Turned Our Logo Into a Living Character: The Case for a Brand Mascot URL: https://www.synapticlabs.ai/blogs/brand-mascot-logo-to-living-character Author: Professor Synapse Published: 2026-10-13 Source: Synaptic Labs, https://www.synapticlabs.ai/blogs/brand-mascot-logo-to-living-character --- > "A logo tells people who you are. A character can show them." _Professor Synapse_ Until September 2026, Professor Synapse, our AI avatar, was mostly a picture. He was a glass flask with a circuit-board brain and three bubbles rising from his neck, sitting between the words SYNAPTIC and LABS in our logo. He looked friendly, and apart from a short animated logo where his flask fills with liquid, he stayed still. Today he waves. He ponders with his thumb and finger on his chin. He sings opera with a hand on his heart, does push-ups, curls a dumbbell, and pushes a plug into its socket with a little spark. When a storm cloud gathers over his head, a bolt hits his bulb, he goes rigid and electrified, and then he shakes it off, sooty and dazed. ![Animated clip of Professor Synapse, a teal flask with a circuit-brain face, struck by lightning: a storm cloud gathers, a bolt hits his bulb, he goes rigid and sooty, then shakes it off](https://cms.synapticlabs.ai/api/media/file/9f4e26391f8abeda6d9de59e58d0014d94a75839f627ad7991531a9db0d2548a.webp) This series is about how a small team turned a logo into a working animated character, with AI agents doing much of the build and people making the calls. It is written for leaders who want to know what it took, what a character could do for their own organization, and when it is worth the effort. This first post covers the why. ## Why a brand character earns its place A brand mascot is a character that stands for an organization and shows up wherever it does: in videos, on the website, in social posts and on slides. It turns a mark people glance at into someone they can follow. We wanted one for three reasons. **People warm to a face.** A flask with eyes, brows and hands can look curious, pleased or confused. A logo can't. When the Professor points at a checklist or shrugs at a bad idea, viewers read the feeling before they read the words. **One character carries across everything.** The same Professor is built for short videos and explainers, and he already appears on our website and in our logo. Whatever the format, people see the same guide. **He stays consistent because he is built once.** Every video uses the same proportions, the same teal, the same ring eyes and the same bubbles. Nobody redraws him for each new project, so nobody can redraw him slightly wrong. For a nonprofit or small business, that last point matters most. Consistency usually costs money: an illustrator on retainer, a style guide, someone checking every new drawing. With a character built once, consistency is the default instead of an extra. ### A brand mascot, or in our case, an AI avatar We call the Professor our AI avatar. He does a brand mascot's job, standing for Synaptic Labs wherever we show up, and AI agents did much of the work of building him. He doesn't come from an AI avatar generator, though. Every appearance comes from one model we built in code. ## He's code, not clips: how our brand mascot animation works The decision that shaped everything else: the Professor is not a folder of video clips or drawings. He is a model built in code (animators call this a rig), and every appearance we make comes from that one model. As of September 2026, he has 63 ready-made moves and 6 emotions, 38 hand poses, and 13 effects (a storm cloud, a lightning bolt, confetti, steam, a lightbulb, a spark and more). He has props too: a clipboard whose page takes any text, a dumbbell, a plug and socket, headphones, a Bunsen burner and a crate to hop onto. ![Contact sheet of Professor Synapse, a teal flask with a circuit-brain face, struck by lightning: a cloud gathers, a bolt hits, one pale flash, he goes rigid, then sooty and dazed, and returns to rest](https://cms.synapticlabs.ai/api/media/file/20155008527ed53060d534a2948fd588e398f5510541648a5a96406ef7a97970.png) The lightning strike, beat by beat: the cloud gathers, the bolt hits, one flash, he goes rigid and electrified, then sooty, smoking and dazed before he shakes it off and returns to rest. Each move starts and ends at the same resting pose, so moves chain together like words in a sentence. A wave, then a point at a chart, then a thumbs-up. A new video is a new sentence built from words we already have. It doesn't need a new animation budget. ![Contact sheet of three Professor Synapse moves, tapping fingers, ponder and sing (opera), each at key frames in the wide format and the vertical 9:16 format](https://cms.synapticlabs.ai/api/media/file/037e73216189f5f3e0ec6630e72f3229c43ff5f934019db62f4364c5446abb3d.png) A contact sheet from a review round: tapping fingers, pondering and singing, each caught at key frames in the wide format and in the vertical format for shorts. > **Under the hood:** the Professor is built with Remotion, a tool that makes video from code, and three.js for the 3D, and he is rendered frame by frame. One design file holds every design decision: colours, proportions, line weights and motion rules. Change a setting there and it changes in every video. ## Brand mascot design: why we left flat 2D The first version was 2D. On September 24 we rebuilt the flask from our logo as a drawing made entirely in code. It was traced from the brand master file so his resting pose matched the logo almost exactly: the average colour difference was about 2 out of 255. He could fill with liquid, grow his circuit brain, look around, blink and wave with floating emoji-style hands. The next day we started on a 3D version. Judging by what the 3D version added, flat 2D had hit four walls. 1. **Turning.** A flat drawing faces one way. The 3D Professor can nod toward the camera, turn his body and spin a full circle when he dances. 2. **Depth.** In 3D a hand sits somewhere in space. It can come in front of his bulb, reach toward the viewer, or drum its fingers on a table just in front of him. 3. **Holding things.** The 2D hands were a set of flat drawings. Changing a pose meant swapping one drawing for another partway through the move, hidden by a quick pinch. A drawing can't wrap its fingers round a clipboard. The 3D hand bends joint by joint, so his fingers curl round the clipboard's edge and close round a dumbbell's bar. 4. **Glass and liquid.** He is a flask, so he should look like one. In 3D his glass is see-through and his liquid moves. The surface stays level when he tilts, piles up when he slides sideways, and bows when he hops. That motion is simulated from how his body moves, so every new move gets it without extra work. ![Contact sheet of Professor Synapse with props: holding a purple prompt-checklist clipboard, curling a dumbbell, and wearing headphones with music-note bubbles, in wide and vertical frames](https://cms.synapticlabs.ai/api/media/file/bb9fcd2e0efaa2d60dfbde43759023375e1feed6c4a12047ab50f251db7b2699.png) Props in the 3D rig: fingers wrapped round a clipboard whose checklist ticks itself off, a dumbbell curled on the arc of an imagined arm, and headphones with music-note bubbles. > **Under the hood:** the hand's bone lengths come from published X-ray measurements of 66 adult hands, and each pose is solved so fingers touch but never pass through each other. It is a cartoon hand with a real hand's proportions. ## Why not just use AI video or image generators? It's a fair question from an AI company, and we did try. We spent a lot of time on AI image and video generators, used up all our credits, and got results that were less than mediocre. It was a waste of time and energy. Here is why they fell short for a mascot, and why we chose to build instead. **They drift off-model.** A generator redraws the character every time. From clip to clip he comes out a little different: another shade of teal, a new flask shape, eyes in a new place. For a mascot, looking the same every time is the whole job. **They garble hands, text and logos.** These are the details generators still get wrong most often. The Professor's clipboard shows whatever text we type, and it stays sharp and readable on a phone screen. **They can't hit a frame.** In a voice short he needs to react on a particular word. Our rig works frame by frame at 30 frames a second: his talking follows the loudness of the voice track, the captions arrive word by word, and a sound effect lands on the frame we pick. A generated clip gives you a take, and you edit around it. **They can't be re-posed or edited predictably.** When a move reads wrong in review, we change one thing and render again. The finger-tapping move first read as a hand hanging off the side of the flask, so we reworked only the hand: we turned it to face the camera, moved it in close and changed how the fingers lift. The rest of the move stayed as it was. With a generator you re-roll the whole clip and hope. So the rig gives us three things a generator doesn't: consistency, control and reuse. We still use AI image tools where they fit, such as one-off blog art. Our rule of thumb: if it has to look the same every time, we build it. If it's a one-off, generating can be the right call. ## What he's done so far - **Our logo.** In the 2026 brand refresh, the logo's flask and our circle icon were rendered from the same 3D model. The character built from the logo became the logo. - **Website poses.** Sixteen poses, among them wave, point, thumbs-up, think and call-me, are rendered from the rig in one shared framing, so any two can swap on a page without him changing size. - **Voice shorts.** Joseph Rosenbaum, our founder, records his voice, and the Professor talks along, reacts on the right words, and shares the screen with captions and on-screen templates. Watch for them on our channels, starting with the short animation made for this post. - **A template gallery.** Twelve gallery entries cover the six short templates and their main variants (a hook, an explainer, a list, a comparison, a screen walkthrough and a chat prompt), plus data templates for stats, before and after, and charts. ## Honest trade-offs A living character isn't free, and it isn't always the right answer. **The up-front build is real.** It took us a few weeks to build the video infrastructure itself, putting Remotion, three.js, ElevenLabs and the rest together, and then about a week for the Professor on top of it. Behind him are a design specification, a hand rig, a liquid simulation, and review rounds on each new batch of moves. That work pays off only if he appears often. Build a character when you expect to use it every week, not every year. **Sometimes a simple illustration is enough.** If you need a character for one poster, a report cover or a handful of stills, commission an illustration. A rig earns its keep in motion and in repetition. **A character needs guardrails.** He is a brand asset, so he has rules. Ours are simple. Teal is his colour, and his alone: his props use the other brand colours. Nobody recolours, restyles or redraws him. We never put a coloured blob behind him. The rules protect the thing that makes a brand mascot useful, which is that people recognize him at a glance. ## Creating a brand mascot for your organization If you're wondering whether a character could help your organization, start with five questions. 1. **Is there a character hiding in your logo or your story already?** Ours was sitting in the middle of our wordmark. 2. **How often will they appear?** Weekly videos and a monthly newsletter justify a build. An annual report may not. 3. **What will they do?** Explain, guide, react, celebrate. Write the list, and you have the first moves. 4. **Who owns the guardrails?** Someone has to say "that's not him" when a draft goes off-model. 5. **What's the smallest useful version?** A set of consistent poses can come first. Motion can follow. This is how we approach any AI project: map the work, train the team, build what fits. A brand character belongs in the build-what-fits column only when the map shows the work repeats. Next in the series: how the Professor learned to move, and why a library of reusable moves makes each new video cheaper than the last. **Want a character like the Professor for your organization, or animated explainers built this way? Let's talk. [Book an AI Opportunities Discussion](https://www.synapticlabs.ai/book-a-call).** ## Frequently Asked Questions ### What is a brand mascot? A brand mascot is a character that represents an organization and appears across its videos, website, social posts and materials. It gives people a face to recognize and follow, where a logo only gives them a mark. ### What is an AI avatar? An AI avatar is a digital character that represents a person or an organization. Ours, Professor Synapse, is a brand mascot built as a 3D model in code, so he looks the same in every video and can talk along with a recorded voice. ### How do you create a brand mascot? Start from something your organization already has, such as a shape in your logo, then define how the character looks, moves and behaves in a short design guide. Test it in a few real pieces of content before investing in animation. We built ours as a 3D model in code so every appearance comes from the same source. ### How do you integrate a brand mascot into marketing? Give the mascot a job in the content you already make: opening a video, pointing at the key idea, reacting to a result, or inviting people to book a call. Brand mascot marketing works best when the character appears consistently in the same role, so people learn what it means when they see it. ### What makes a brand mascot memorable and effective? Consistency and personality. A mascot is memorable when it looks the same every time and shows clear feelings, such as curiosity, delight or confusion. It is effective when it serves the message instead of competing with it. ### How can a brand mascot make your business more personable? A character can say things a logo can't: it can welcome people, admit something is tricky, or celebrate a small win. That gives a small organization a warm, recognizable voice without putting one staff member in front of the camera for everything. ### Can AI image or video generators make a consistent brand mascot? They can make a one-off image quickly, but they tend to redraw the character slightly differently each time, and they struggle with hands, text and logos. For a mascot that must look identical across many videos and react on a precise word, a character built once and reused gives more consistency and control.