Skip to content

OpenRAG is a comprehensive, single package Retrieval-Augmented Generation platform built on Langflow, Docling, and Opensearch.

Python
3.8 KB
Raw Blame
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>OpenRAG Sample Knowledge Base</title>
</head>
<body>

<header>
    <h1>OpenRAG Sample Knowledge Base</h1>
    <p>
        This document is intended for testing document ingestion,
        semantic search, chunking, metadata extraction, and chatbot retrieval.
    </p>
</header>

<hr>

<section id="company">
    <h2>Company Overview</h2>

    <p>
        Contoso AI Solutions develops enterprise software for document search,
        Retrieval-Augmented Generation (RAG), AI assistants, and workflow automation.
    </p>

    <p>
        Headquarters: Seattle, Washington<br>
        Founded: 2021<br>
        Employees: 180
    </p>
</section>

<section id="products">
    <h2>Products</h2>

    <ul>
        <li>OpenRAG Enterprise</li>
        <li>Document Search API</li>
        <li>AI Knowledge Assistant</li>
        <li>Workflow Automation Platform</li>
    </ul>
</section>

<section id="features">
    <h2>Platform Features</h2>

    <table border="1" cellpadding="8">
        <thead>
            <tr>
                <th>Feature</th>
                <th>Description</th>
                <th>Status</th>
            </tr>
        </thead>
        <tbody>
            <tr>
                <td>Document Upload</td>
                <td>Supports PDF, DOCX, HTML, TXT, and Markdown.</td>
                <td>Available</td>
            </tr>
            <tr>
                <td>Semantic Search</td>
                <td>Vector search using embeddings.</td>
                <td>Available</td>
            </tr>
            <tr>
                <td>Chat with Documents</td>
                <td>Answers questions using indexed documents.</td>
                <td>Available</td>
            </tr>
            <tr>
                <td>Multilingual Support</td>
                <td>Supports more than 40 languages.</td>
                <td>Beta</td>
            </tr>
        </tbody>
    </table>
</section>

<section id="installation">
    <h2>Installation</h2>

<pre><code>
git clone https://github.com/example/openrag-demo.git

cd openrag-demo

pip install -r requirements.txt

python app.py
</code></pre>

</section>

<section id="configuration">
    <h2>Configuration</h2>

    <p>
        Configure the application using environment variables:
    </p>

<pre><code>
OPENAI_API_KEY=your_api_key
VECTOR_DB=qdrant
CHUNK_SIZE=1000
CHUNK_OVERLAP=200
EMBEDDING_MODEL=text-embedding-3-large
</code></pre>

</section>

<section id="workflow">
    <h2>Document Processing Workflow</h2>

    <ol>
        <li>Upload a document.</li>
        <li>Extract text.</li>
        <li>Split into chunks.</li>
        <li>Create embeddings.</li>
        <li>Store vectors.</li>
        <li>Perform semantic retrieval.</li>
        <li>Generate answers using the retrieved context.</li>
    </ol>
</section>

<section id="faq">
    <h2>Frequently Asked Questions</h2>

    <h3>What file formats are supported?</h3>
    <p>
        PDF, HTML, DOCX, TXT, Markdown, CSV, and JSON.
    </p>

    <h3>Which embedding model is used?</h3>
    <p>
        The default embedding model is
        <strong>text-embedding-3-large</strong>.
    </p>

    <h3>Can I upload multiple documents?</h3>
    <p>
        Yes. Documents are indexed independently while sharing the same
        searchable vector database.
    </p>

    <h3>Does the system support citations?</h3>
    <p>
        Yes. Retrieved chunks can be returned with source references.
    </p>
</section>

<section id="contact">
    <h2>Support</h2>

    <p>
        Email: [email protected]
    </p>

    <p>
        Documentation:
        <a href="https://example.com/docs">
            https://example.com/docs
        </a>
    </p>
</section>

<footer>
    <hr>
    <p>
        End of OpenRAG Sample Document.
    </p>
</footer>

</body>
</html>