OnBrand by SlideSpeak

Product Introduction

1. Definition: OnBrand by SlideSpeak is an AI-powered brand management platform and Model Context Protocol (MCP) server. It functions as a centralized design context repository, providing structured brand assets and rules directly to AI agents like Claude, Codex, and ChatGPT.
2. Core Value Proposition: The platform's primary purpose is to enforce brand consistency and eliminate generic, off-brand output from generative AI tools. It solves the "brand amnesia" problem by giving AI agents access to a company's true logos, colors, typography, approved slide layouts, and imagery, enabling them to generate on-brand presentations, infographics, and design assets autonomously.

Main Features

1. Brand Asset Library & Token Management: This feature involves ingesting and structuring a company's entire visual identity system. Users upload logos, color palettes, font files, icon sets, and imagery guidelines, which the system processes into machine-readable design tokens (e.g., exact hex codes, font names, layout HTML/CSS). It can automatically generate a design.md brand guide from existing websites, PDFs, or PowerPoint templates, creating a single source of truth for AI consumption.
2. Model Context Protocol (MCP) Integration: The core technical innovation is the exposure of the brand library via a standard MCP server endpoint. Any MCP-compliant client (like Claude Code, Codex, or custom agents) connects to a single API. Through standardized tools (get_brand, get_layout, check_brand), the agent can query for color tokens, fetch approved slide layout HTML/CSS, and validate that generated content adheres to brand rules, all in real-time.
3. Governed Brand Hub Dashboard: This is the administrative control panel where brand managers govern access. It provides features for version control of assets, role-based access control (RBAC) for teams and regions, audit logs to track which agents used which assets, and permissions management. It scales to support multiple sub-brands and locales under one account, with enterprise security features like SSO and SCIM.

Problems Solved

1. Pain Point: The core problem addressed is inconsistent AI-generated branding. AI models, lacking inherent brand knowledge, produce outputs with incorrect logos, wrong color palettes, unfamiliar fonts, and generic layouts, forcing human teams to manually redesign assets. This creates significant inefficiency and brand drift at scale.
2. Target Audience: Key users include Enterprise Brand Managers, Marketing Operations Teams, Creative Directors, Consulting Firms producing client decks, and IT/Engineering Teams building custom AI agent pipelines. It is designed for organizations where brand consistency is critical and AI tools are being adopted at scale.
3. Use Cases: Essential scenarios include generating quarterly sales decks, creating investor pitch decks, producing client deliverables, building marketing one-pagers, and producing internal communications—all directly within AI workflows while maintaining strict adherence to brand guidelines.

Unique Advantages

1. Differentiation: Unlike generic AI presentation makers (like Gamma or generic templates) or standard DAM systems, OnBrand doesn't just store assets; it actively injects brand context into the AI generation process via MCP. It is agent-agnostic and works at the API level, integrating into developer and enterprise workflows rather than being a standalone app.
2. Key Innovation: The key innovation is leveraging the Model Context Protocol (MCP) as a standard interface for AI agents to access design systems. This moves beyond simple prompt engineering or style transfer, providing a governed, versioned, and auditable pipeline for brand-compliant AI content generation. The check_brand tool for real-time validation of AI output is a unique technical capability.

Frequently Asked Questions (FAQ)

1. How does OnBrand integrate with ChatGPT and Claude? OnBrand integrates via the Model Context Protocol (MCP). Users configure their AI agent clients (like Claude Desktop or ChatGPT's custom GPTs) to connect to their OnBrand MCP server endpoint. The agent then gains access to tools for fetching brand assets and validating layouts, enabling on-brand generation directly within the chat interface.
2. What specific AI tools and platforms is SlideSpeak OnBrand compatible with? It is compatible with any MCP-compliant client. Explicitly supported platforms include Claude Code, Codex, Cursor, and custom agents built on the protocol. It works with AI presentation generators within these environments.
3. Can I import my existing brand guidelines from a PDF or PowerPoint? Yes. OnBrand's ingestion system can analyze existing brand guideline PDFs, PowerPoint templates, and websites to automatically extract and structure logos, color palettes, fonts, and slide masters into its managed brand kit.
4. What security and governance features does it offer for enterprise use? For enterprise deployment, OnBrand provides Role-Based Access Control (RBAC), SSO (Single Sign-On) and SCIM provisioning, a full audit trail of asset usage by AI agents, version control with rollbacks, and the ability to manage multiple brands or sub-brands with isolated contexts under one account.