Chat.sh

Product Introduction

1. Definition: Chat.sh is a self-hosted, AI-powered knowledge base and help center platform. Technically, it is a SaaS application that provides a content management system (CMS) for documentation, an AI-powered semantic search engine, and a deployment engine that serves the compiled help center from a designated folder on a user's existing web server or custom domain.
2. Core Value Proposition: Chat.sh exists to replace traditional, subscription-based help center software with a one-time-purchase solution that provides intelligent, context-aware answers directly on a company's own domain. Its primary value is delivering AI search that answers questions instead of just returning links, while eliminating recurring monthly fees.

Main Features

1. AI-Powered Semantic Search: The core feature is a search engine that uses a language model to understand the intent behind a user's natural language query. It retrieves relevant passages from the knowledge base, synthesizes an answer, and cites the source pages. This contrasts with traditional keyword-based search. It works by embedding content into vectors and using retrieval-augmented generation (RAG) to provide accurate, sourced answers.
2. Self-Hosted Folder Deployment: Unlike platforms that force a subdomain (e.g., help.yoursite.com), Chat.sh allows deployment to a folder on an existing website (e.g., yoursite.com/help or yoursite.com/guide). This is achieved by generating static HTML, CSS, and JavaScript files that users upload to their web server, maintaining brand consistency and simplifying analytics.
3. LLM-Optimized Knowledge Base Export: Every article in the knowledge base is available as raw Markdown. Users can copy individual pages or the entire documentation set for use with AI assistants like ChatGPT and Claude. The platform also supports an llms.txt file, a standardized endpoint for AI agents to crawl and ingest documentation context directly.
4. Unified Knowledge Base for Multiple Channels: Content is created once in the central knowledge base and powers three distinct channels: the public help center website, an upcoming chat messenger widget for in-app support, and the AI-agent-ready Markdown exports. This ensures answer consistency across all customer support touchpoints.
5. Lifetime License with Predictable Pricing: Chat.sh operates on a one-time payment model (Lifetime Deal) rather than a subscription. The license includes all current and future core features (help center, messenger, support inbox) for a single upfront fee, with pricing based on AI credit and page limits per workspace.

Problems Solved

1. Pain Point: Ineffective help center search that returns irrelevant lists of article titles instead of direct answers, leading to poor user experience and increased support ticket volume.
2. Pain Point: Loss of brand control and SEO value when help centers are hosted on third-party subdomains, separate from the main company website.
3. Pain Point: Escalating and unpredictable costs associated with monthly subscriptions for help desk and knowledge base software.
4. Target Audience: Founders and product teams at SaaS companies and startups; developers and technical leads responsible for developer documentation; customer support managers seeking to defray ticket volume with scalable self-service.
5. Use Cases: A SaaS company deploying a public help center at app.yoursite.com/docs; a developer tools company providing API documentation optimized for AI ingestion via llms.txt; a bootstrapped business replacing Intercom or Zendesk Guide to cut recurring operational costs.

Unique Advantages

1. Differentiation vs. Traditional Help Desks (Intercom, Zendesk): Chat.sh focuses exclusively on the knowledge base and AI search, offering a superior answer engine without the bloat of full CRM suites. It competes on superior search technology and a one-time cost versus perpetual subscriptions.
2. Differentiation vs. Modern AI Help Desks (HelpKit, Mintlify): While others offer AI search, Chat.sh's key differentiator is the self-hosted folder deployment. This gives users complete control over hosting, performance, and integration with their existing site infrastructure, which is often a requirement for security-conscious or branding-focused teams.
3. Key Innovation: The combination of folder-based deployment with advanced RAG-based AI search. This solves both the technical hosting problem and the user experience problem in a single, purchase-once product. The llms.txt support further future-proofs the platform for the age of AI agents.

Frequently Asked Questions (FAQ)

1. How does Chat.sh's AI search work compared to a normal help center? Traditional help center search acts like a keyword index, matching words in your query to words in article titles. Chat.sh's AI search uses a language model to understand the meaning of the question, finds relevant information across all your documentation, and writes a concise answer while citing its sources, much like a human support agent would.
2. Can I really host the Chat.sh help center on my own website without a subdomain? Yes. Chat.sh generates static files that you can upload to a folder (like /help or /docs) on your existing web server (e.g., NGINX, Apache, Vercel, Netlify). This means your help center lives at yoursite.com/help, sharing your domain's SSL certificate, analytics, and SEO authority.
3. What is included in the Chat.sh lifetime deal? The lifetime deal is a one-time payment for a perpetual license for one workspace. It includes the AI help center, the upcoming chat messenger widget, and the future support inbox feature. The license is defined by limits (e.g., 1,000 pages, 1,000 AI answers/month) which are permanent for that workspace, with the option to purchase additional AI credit packs.
4. What is llms.txt and why is it important for AI? llms.txt is a proposed standard file (like robots.txt) that tells AI agents and web crawlers which parts of your site are allowed to be indexed for training. By supporting this, Chat.sh ensures your documentation is structured and accessible for future AI tools, making your knowledge base ready for the next generation of search and support automation.
5. Is Chat.sh suitable for large, complex documentation sets? Chat.sh is designed for scale, supporting up to 1,000 pages (including articles, web links, and files) in its lifetime tier. Its AI search is particularly effective for large knowledge bases where users struggle to find answers using traditional search. For context, a page is counted as ~2,500 characters of text or one PDF page.