Mark (DYM)
The Artist of Presence
Mark (DYM) is your palette first visual and media orchestration system. It uses Palettes and channel presets to generate, transform, and deliver on-brand images into the tools you already use. Where visuals remember the rules.
Visual and media orchestration system for on-brand AI images
Mark (DYM)
The Artist of Presence
Mark (DYM) is your palette first visual and media orchestration system. It uses Palettes and channel presets to generate, transform, and deliver on-brand images into the tools you already use. Where visuals remember the rules.
Presence becomes legible when form is given rules.
Clarity begins with naming what the system actually holds.
What it is
For everyone
Mark (DYM) is your visual and media orchestration system for on-brand AI images. Instead of asking a model to improvise every banner, thumbnail, or social creative, you keep visual logic in Palettes.
A Palette is your brand control pack: colors, typography tokens, logo assets, watermark policy, disallowed subjects, and channel presets.
You send a request with a prompt or structured JSON, optionally with a reference image. Mark:
- Assembles a Visual Prompt Package from Palette, channel preset, and request-time overrides
- Calls the image capability for 1 to N outputs
- Runs transforms such as resize, crop, background remove, format export, and watermark
- Stores originals and variants with provenance
- Delivers assets by webhook or signed URL
The result is a brand-aware image generation API with Palettes and webhook-first delivery that teams can trust and automate, instead of one-off prompts and manual exports.
Technical view
Mark (DYM) models a fixed visual pipeline:
Organization → Project → Palette → Asset → Render → Transform → Deliver
Palette holds brand controls and channel presets.
Render assembles the Visual Prompt Package (Palette plus preset plus overrides), calls the managed image capability, and returns 1 to N outputs.
Transform runs post-render operations such as resize, crop-to-channel, background remove, format export, and watermark.
Safety and policy run before delivery. Unsafe assets are quarantined and not sent.
Deliver hands off assets and metadata via webhooks or signed URLs, emitting render.status events such as queued, rendered, delivered, and errored.
Because Mark is standalone and API-first, it can consume structured context from your systems and return versioned assets with provenance and auditability to your automations, applications, or DAM and CDN layers.
When visual rules stay implicit, presence falls apart.
The problem it solves
Off-brand generation
AI images ignore color, logo, watermark, or basic brand rules.
Unrepeatable prompts
Prompts cannot be reproduced, so approvals have no trace or single source of truth.
Manual resizing and exports
Teams resize and export every ratio and format by hand for each channel.
Missing provenance
Nobody can answer which Palette, which model, or which rules created a given image.
Scattered delivery
Assets travel through chats and links instead of webhook-first delivery with render.status events.
Safety as an optional step
Watermarks and disallowed subjects depend on memory instead of enforced, non-bypassable policy.
When visual context frays, presence dissolves. Structure lets intention return to every asset.
Every layer is deliberate. Nothing is improvised.
How it works
Create Project and Palette
Create a Project per brand or client and define one or more Palettes with colors, typography tokens, logos, channel presets, disallowed subjects, and watermark policy.
Brand controls move into versioned Palettes, so on-brand images become the default instead of an exception.
Prepare input
Send a prompt or structured JSON payload, optionally with reference images and request-time overrides for size, ratio, crop, or format.
Teams can start with simple text prompts and grow into structured context without changing the core pipeline.
Assemble the Visual Prompt Package
Mark merges the Palette at a specific version, the channel preset, and request-time overrides into a Visual Prompt Package that actually goes to the image capability.
Assembly is deterministic and inspectable, so you always know which rules shaped each render.
Route and Render (images in MVP)
Mark selects a managed image capability, executes the Render job with the Visual Prompt Package, and produces one or more outputs while capturing model, parameters, and timing.
Image generation becomes predictable, with explicit error codes such as DYM-RENDER-MODEL_UNAVAILABLE instead of silent failures.
Transform and safeguard
Post-render, Mark can resize and crop to channel presets, remove background, change format, apply watermark policy, and run safety scans that enforce disallowed subjects.
You receive channel-ready variants and enforced visual policy in one call instead of layers of manual work.
Store, deliver, observe
Mark stores originals and variants per Project with metadata and provenance, delivers assets via webhooks or signed URLs with storage handoff, and emits render.status events with observability data.
Assets become first-class, traceable objects, and operations can debug issues with clear signals rather than guesswork.
Reasoning in visuals is assembled step by step. Order becomes quiet architecture.
Structure becomes advantage when presence is no longer left to chance.
Why this product
Palette-first pipeline — Brand controls precede generation. Palettes and channel presets define what is allowed before any pixels exist, so creative freedom lives inside clear constraints.
Versioned assets with provenance and auditability — Every Asset can answer which Palette, which model, which rules, which request, and which version of the Visual Prompt Package created it.
Webhook-first media delivery with storage handoff — Delivery is designed around webhooks, signed URLs, and storage handoff so images fit into your existing stack instead of creating a new silo or public CDN.
Standalone and API-first — Prompt or JSON in, asset out. Mark works alone, and when other YounndAI products are present it can consume their structured context without depending on them.
Fixed stages and clear error codes — Palette, Render, Transform, Deliver, and safety checks follow a fixed Palette → Render → Transform → Deliver pipeline with explicit error codes. Fewer meetings, faster debugging.
Governance built in — Roles, signed Palettes, scoped secrets, safety, and logs are part of the core design rather than retrofitted extras.
When Palette, provenance, and delivery move together, teams stop guessing and start trusting their visual stack.
Different roles, one place to hold the visual rules.
Who it's for
Treat Mark as the visual and media orchestration layer in front of image models so on-brand assets become the default output of AI features.
Generate square, vertical, and landscape assets that already match brand look and format, without last minute exports.
Use a single Palette-based visual layer for product shots, feature launches, and social creatives so small teams look consistent before they can hire full creative departments.
Give each client its own Project and Palettes so brand controls, watermarks, and delivery endpoints stay separated, auditable, and easy to manage.
Rely on provenance, signed Palettes, safety scans, and audit logs to keep AI visuals compliant without blocking teams.
Use Mark as an image generation API inside n8n, Make, Zapier, or custom webhooks to render, transform, and deliver assets inside end-to-end workflows.
Systems earn trust when they solve named visual moments.
Real-world uses
Scenario
You run campaigns that need square and vertical visuals with the same look.
Action
POST prompt plus palette_id plus channel preset. Mark renders size-locked images, applies watermark, and delivers signed URLs by webhook.
Outcome
Feeds stay visually consistent across teams and tools with less manual export work.
Structure keeps visual scenarios repeatable across teams, systems, and time.
A visual layer only matters if it can be called from anywhere.
Integrations
Connect what you already use
If it speaks HTTP, Mark can render and hand it off.
Clarity in visuals also means knowing who can do what.
Security & Governance
Security and governance in Mark (DYM) are defined up front. Roles, scoped secrets, and audit logs give governed teams a clear view of every Render, Transform, and Deliver.
Access Controls
- Members see nothing until they are invited.
- Organization admins create Projects.
- Project admins control deletions and other destructive actions.
- Secrets are scoped per organization, project, and integration so each key has only the access it needs.
- Palettes can be signed with visible authors and trust levels so teams know who defined which rules.
Behaviors
- Watermark policy in a Palette is non-bypassable when enabled.
- Audit logs record Render, Transform, and Deliver events with sensitive fields redacted.
- Error codes such as DYM-DELIVER-WEBHOOK_4XX or DYM-VALIDATION-SCHEMA_FAIL make debugging predictable instead of opaque.
Layers arrive in sequence so structure can stay honest.
Product status
Now
- Organizations and Projects
- Palettes with brand controls and static channel presets
- Image renders with 1 to N outputs
- Image transform API for resize, crop, background remove, format export, and watermark
- Light storage per Project for originals and variants
- Webhook and URL delivery with signed links
- Provenance header on every Asset
- Basic safety scan and quarantine
- Render Linter and Schema Validation, plus a Debug View of the Visual Prompt Package
- Standard error codes for Render, Transform, and Deliver
- Observability for render IDs, latency breakdown, and Compute Units
Next
- Sequences for short video and motion clips
- Bulk and batch operations and multi-ratio packs
- External model connectors so you can bring your own image or video models
- Palette registry and sharing for reusable visual packs
- Golden Renders and Asset diffing for quality assurance
- Policy overrides per tenant or client
Questions reveal where structure must speak more clearly.
FAQ
Every creative system begins with a question. Here are ours.
YounndAI (pronounced 'yoon-dye') is the philosophy and architecture of human first intelligence that unifies all Elements and Systems. It is a human-first way of building AI that follows four principles: Discipline with Flow, Human before Machine, Structure before Scale, and Continuity before Chaos. It says intelligence should be structured, human-first, and continuous, not improvised or extractive. It is the architecture beneath the products, not a product or platform by itself. YounndAI means you and AI, unified.
Clarity is recursive. Ask again when the system grows.
From the first stroke of color, presence can become coherent instead of improvised.
Ready to put Mark in your visual stack?
Define your Palettes once. Mark (DYM) renders, transforms, and delivers on-brand images through a brand-aware image generation API with Palettes and webhook-first media delivery.
Mark is one stroke in a larger form.
Products connect cleanly when they share the same structure.
The architecture beneath the products explains the shape they take.
YounndAI is the philosophy that says intelligence should be structured, human-first, and continuous. Every product, including Mark (DYM), follows four principles.
Discipline with Flow
Structure and intuition stay in balance. Palettes, pipelines, and error codes give visuals a clear frame while teams stay free to explore styles, experiments, and campaigns.
Human before Machine
Humans keep final control. Transparent metering, user-controlled data, and clear safety rules mean that keys, context, and costs belong to the Organization, not the platform.
Structure before Scale
Complexity is earned through clarity. A fixed Palette → Render → Transform → Deliver pipeline, explicit entities, and provenance headers keep visual work ordered so growth adds volume without adding confusion.
Continuity before Chaos
Memory sustains meaning. History, logs, and reproducible Assets ensure that visual intent does not erode as teams iterate on Palettes, models, and campaigns.
Define. Build. Remember. Harmonize. Harmony above all.