One brand profile. Everything else follows.
Three groups do the work: a Foundation that remembers who you are, five creation engines that read it, and the intelligence layer that runs it all, with your permission.
The part that remembers you.
Brand Persona
Live31 fields across Audience, Brand Core, Personality & Voice, Visual Direction, Market & Differentiation, and Constraints & Context, saved and scored independently.
Explore →Brand Kit
LivePalette, typography, logo direction, strategy and visual assets, regenerate any module, keep what you love.
Explore →Quality Engine
LiveCompleteness scoring and weakness alerts, it tells you where your brand is thin, and helps you fix it.
Explore →
Five engines, one voice.
Branding
LiveA full brand kit from your persona.
Explore →AI Twins
LiveAn AI that sounds, writes, and looks like you.
Explore →Carousels
Rolling outSwipeable posts with hook, body, CTA.
Explore →Static Posts
Rolling outSingle-image posts, on brand.
Explore →Stories
Rolling outVertical frames sized per platform.
Explore →
Described once. Read everywhere.
The layer that runs it, and asks first.
Repic Assistant
LiveIt doesn't just answer, it does the work. Impactful changes wait behind an Allow/Disallow card.
Explore →Voice Chat
LiveTalk to your assistant out loud, and hear it answer in your own cloned voice if you choose.
Explore →Canva integration
Open any asset in Canva, edit it, and it round-trips back into your kit.
Explore →
Precision tools that already know your brand.
Transcription, image and video generation, upscaling, virtual try-on and Instagram insights, all inside the studio, all grounded in the same brand profile.
3 min read · 5 sections
How the six engines actually fit together
Most AI content tools are a template with a text box in front of it. You pick a layout, you type a topic, and the tool guesses at everything else: the tone, the colours, the audience, the point. Guess again next week and you get a different guess. That is why a feed built out of single-purpose generators reads like it was assembled by five people who never spoke to each other.
Keep reading
Repic is built the other way round. You describe your brand once, in real depth, and that description becomes a durable object every engine reads before it writes a single word. Every engine reads the same source of truth: your Brand Persona (the 31-field description of who you are, who you serve and how you sound) and the Brand Kit generated from it (palette, typography, logo direction, voice and strategy).
The practical consequence is that consistency is not something you police afterwards. A carousel and a story generated three weeks apart share a palette because they read the same palette, not because you remembered to set it twice. When your positioning changes, you change it in one place and the Quality Engine tells you which parts of your kit are now out of date.
Step by step, with nothing skipped
- Foundation
Branding writes your source of truth
The Branding engine turns your answers into a Brand Persona and a Brand Kit. This is the only step where you talk about yourself; everything after it reads what you said.
- Identity
Your Brand Kit becomes reusable
Palette, type pairing, logo direction, voice rules and strategy, generated together so they agree with each other rather than being five separate decisions.
- Likeness
AI Twins add you, with consent
A voice, writing and look twin, each gated behind explicit consent. Keep as many versions as you like; exactly one is active per modality, so output is never ambiguous about which you it used.
- Output
Five creation engines read the kit
Carousels, static posts, stories, long content and the Assistant. None of them asks you for your colours, because none of them has to.
- Health
The Quality Engine keeps it honest
Completeness scoring per component, plain-language weakness alerts, and staleness tracking when a source field changes underneath generated work.
What this does not do
It is not a scheduler
Repic makes the work and hands it to you finished. Publishing, queues and cross-posting are not what this is.
It is not a template marketplace
There is no library of layouts to browse. The structure comes from your brand and the format, which is the entire point.
It is not a team collaboration suite
Agencies run many client brands as separate sources of truth, but there are no approval chains, comment threads or role hierarchies.
Straight answers
Do I have to fill in the whole brand profile before I can generate anything?
No. Every persona field and kit module is saved and scored independently, so you can generate from a partial brand. The Quality Engine will mark thin components as Underspecified and tell you exactly which two or three fields would lift the output most, rather than blocking you behind a form.
What happens to work I already generated if I change my brand?
Nothing is silently rewritten. Past runs stay exactly as generated, in your history, versioned. What changes is that Repic marks which parts of your kit are now out of date relative to the field you edited, so you know what is worth regenerating instead of guessing.
Can I use Repic for more than one brand?
Yes. Each brand is its own source of truth with its own persona, kit, twins and history. That is how the agency use case works: one client, one brand profile, no cross-contamination between them.
Which engines are live right now?
The status shown on each engine page is read from the live capability registry, not hardcoded, so a page never claims a rolling-out engine is finished. If a page says Rolling out, that is what it means.