NOAN

Product Introduction

1. Definition: NOAN is a verified fact layer API and MCP (Model Context Protocol) server designed for agentic business operations. It functions as a centralized, version-controlled source of truth for company data.
2. Core Value Proposition: It exists to eliminate AI hallucination and data inconsistency by providing a single, approved, and live API endpoint for all business facts—such as pricing, product details, policies, and customer information—ensuring every AI agent, application, and team member operates from the same grounded context.

Main Features

1. Verified & Versioned Fact Layer: NOAN's core is a database of "facts"—company-approved data points. Each fact is versioned, carries an approval history (who approved it and when), and supersedes previous versions. This creates an immutable audit trail. How it works: Facts are managed via a web interface or API, and any connected system performs a simple RESTful GET request to retrieve the current, verified version.
2. Universal API & MCP Integration: The platform exposes its fact layer through a standard REST API and the emerging Model Context Protocol (MCP). This model-agnostic approach means it works with any AI model or agent framework, including Claude, OpenAI GPTs, Llama, and custom agents, as well as applications like Cursor, Replit, and web platforms.
3. Agentic Task Board with Full Audit Trail: NOAN integrates a shared task board where both human team members and AI agents can claim and complete work. Every agent action is logged on the record, providing transparency. This allows for the orchestration of a "fleet" of specialized agents (e.g., for support, sales follow-up) triggered by tags, with all activity visible in a unified Activity feed.
4. Live Data Synchronization ("Change Once, Update Everywhere"): When a core fact is updated in NOAN, every connected endpoint updates instantly. This powers live website pricing pages, ensures all AI agents quote correct information, and syncs internal tools without manual deployment, acting as a headless CMS for operational data.
5. Verity - The Slack-Integrated Agent: NOAN provides a pre-built AI agent, Verity, that operates within Slack. She answers team questions based on verified facts, drafts potential facts from conversations for approval, creates tasks in NOAN, and manages other specialized agents, bringing the fact layer into existing communication workflows.

Problems Solved

1. Pain Point: AI Hallucination and Inconsistent Data in Agentic Workflows. When AI agents scrape or recall information from various documents (PDFs, wikis, chats), they often produce conflicting or outdated answers, leading to errors in customer quotes, support, and marketing.
2. Target Audience: Agentic Startup Teams, AI Product Engineers, Operations Managers, and Founders who are building with AI agents and need a reliable, scalable source of truth. It's also critical for Marketing and Sales Teams requiring accurate, synchronized data across websites, proposals, and CRM tools.
3. Use Cases:
   • Ensuring an AI sales agent always quotes the correct, approved pricing and product specs.
   • Powering a company website where pricing and feature data is rendered live from the fact layer, eliminating static page updates.
   • Providing a single context source for coding assistants (like Claude Code) to build internal tools that use accurate company data.
   • Automating customer onboarding, support follow-ups, and content generation (like changelogs) with agents that pull from verified facts.

Unique Advantages

1. Differentiation: Unlike vector databases or traditional knowledge bases that store "what was said," NOAN mandates a verification and approval workflow to establish "what is true." Unlike using a model's long context window, which is expensive and passive, NOAN actively serves specific, concise facts via API, drastically reducing token costs and improving accuracy.
2. Key Innovation: The combination of a human-in-the-loop verification system with a live-sync API and MCP server. This bridges the gap between human governance (approval, versioning) and machine consumption (low-latency API calls for agents and apps), creating a controllable "company brain" for the AI era.

Frequently Asked Questions (FAQ)

1. How is NOAN different from using ChatGPT with a uploaded document? ChatGPT with document upload provides memory of a static document snapshot. NOAN provides a dynamic, version-controlled, and approved source of truth via API. Changes in NOAN are instantly available to all connected systems, whereas document-based approaches require re-uploading and lack approval trails and live synchronization.

2. What is the Model Context Protocol (MCP) and why does NOAN support it? MCP is a protocol developed by Anthropic to standardize how external data and tools are provided to AI models. By supporting MCP, NOAN can natively and securely connect to Claude and other MCP-compatible agents without complex custom integrations, making the fact layer a first-class context source within the agent's workflow.

3. Can NOAN replace my CRM or CMS? NOAN is designed as a foundational fact layer, not a full-featured CRM or CMS. However, it can power the core data within those systems. For example, it can serve as the headless backend for your pricing page (CMS function) and store verified customer attributes linked to tasks and notes (lightweight CRM function), syncing with your primary tools via API.

4. Is NOAN suitable for non-technical users? Yes, through interfaces like the Verity Slack bot. Team members can ask questions and create tasks in natural language within Slack. The approval and fact management interface is designed for business users, not just developers, allowing product or sales managers to own and verify the facts relevant to their domain.

5. How does NOAN handle data security and permissions? Access to the NOAN API is controlled via bearer tokens. The platform includes permissioning for facts, allowing teams to control who can view or approve certain data blocks. All changes are versioned and auditable, providing a clear record of who changed what and when.