Core Concepts
This page explains the fundamental concepts behind Onsomble. Understanding these will help you get the most out of the platform.
Notebooks
A Notebook is the primary organizational unit in Onsomble. Think of it as a container for related knowledge and conversations.
Sources
Documents, PDFs, and web content that form your knowledge base
Conversations
AI-powered chat sessions grounded in your content
Workspace Files
Additional files and notes you want to keep together
Knowledge Graph
Visual connections between concepts and entities
Create separate notebooks for different projects or topics to keep your knowledge organized.
Sources
Sources are the documents and content that feed your knowledge base. When you add a source to a notebook, Onsomble processes it to enable intelligent search and chat.
How Source Processing Works
Upload or Add Content
Add your source via file upload, URL, or YouTube link.
Content Extraction
Onsomble extracts text and metadata from your source.
Embedding Generation
The content is converted into vector embeddings for semantic search.
Ready for Chat
Your content is now available for RAG-powered conversations.
Supported Source Types
| Type | Formats | Notes |
|---|---|---|
| Documents | PDF, TXT, MD, DOCX | Most common formats supported |
| Web Content | Any URL | Extracts main content from web pages |
| Video | YouTube | Extracts and processes transcripts |
Large files may take longer to process. You’ll see a processing indicator while content is being prepared.
AI Chat
Onsomble’s AI chat uses RAG (Retrieval Augmented Generation) to provide answers grounded in your actual content.
How RAG Works
You Ask a Question
Type your question in natural language.
Retrieval
Onsomble searches your sources to find relevant passages using semantic similarity.
Augmentation
The relevant content is provided as context to the AI model.
Generation
The AI generates a response based on your content, with citations.
Unlike generic AI chat, RAG ensures answers come from your sources. This means more accurate, relevant responses with verifiable citations.
Best Practices for Questions
Be specific:
- “What were the Q3 revenue figures?” vs “Tell me about revenue”
Reference context:
- “According to the research paper, what methodology was used?”
Ask for summaries:
- “Summarize the key findings from the meeting notes”
Audio Summaries
Generate audio versions of your content for on-the-go learning.
Audio generation consumes credits. Check your credit balance before generating long summaries.
Use Cases
- Listen to document summaries during commutes
- Review meeting notes hands-free
- Create audio study materials
Knowledge Graphs
Knowledge Graphs visualize the relationships between concepts, entities, and ideas within your content.
Components
- Nodes - Represent concepts, people, organizations, or topics
- Edges - Show how nodes relate to each other
- Properties - Additional metadata about nodes and relationships
Powered by Neo4j
Onsomble uses Neo4j, a leading graph database, to power knowledge graphs. This enables:
- Fast traversal of relationships
- Complex queries about connections
- Discovery of non-obvious relationships
Knowledge graphs are especially useful for research projects where understanding connections between concepts is important.
Credits
Onsomble uses a credit system for AI operations to ensure fair resource usage.
Credit Costs
| Operation | Credit Cost |
|---|---|
| Chat message | ~1-3 credits |
| Audio generation | ~5-10 credits |
| Source processing | Free |
Credit costs may vary based on the length and complexity of requests. Check your dashboard for current balance and usage.