> ## Documentation Index > Fetch the complete documentation index at: https://fusiondocs.finchai.com/llms.txt > Use this file to discover all available pages before exploring further. # Searching Documents > Find relevant information across your project documents # Searching Documents Finch Fusion provides powerful search capabilities to help you find relevant information across all documents in your project. The platform uses AI-powered vector search to understand the meaning of your queries, not just keyword matching. ## How Search Works Unlike traditional keyword search, Finch Fusion uses semantic search powered by vector embeddings. This means: * Searches understand the **meaning** behind your query * Results are ranked by **relevance**, not just keyword frequency * You can find information even when exact words don't match * Related concepts and synonyms are considered ## Using Search ### Quick Search Use the search functionality to quickly find information across your project: 1. Enter your search query in natural language 2. Results appear ranked by relevance 3. Click a result to view the source document ### Search Through Chat The most common way to search is through the chat panel: 1. Type your question in the chat input 2. The AI searches your documents automatically 3. Answers include citations linking to the source **Example queries:** * "What was the company's revenue in Q4?" * "Find information about supply chain risks" * "Show me details about the CEO's compensation" ## Search Suggestions As you type, the platform may offer search suggestions based on: * Common queries for your document types * Previously successful searches * Key topics extracted from your documents ## Search Results Results include: * **Relevant text passages** from your documents * **Source document** identification * **Relevance score** indicating how well the result matches your query * **Context** around the matching passage ## Tips for Better Results * **Be specific** - "What was Apple's services revenue in Q3 2025?" works better than "revenue" * **Use natural language** - Write queries as you would ask a colleague * **Include context** - Mention the document type, time period, or topic area * **Try different phrasings** - If initial results aren't satisfactory, rephrase your query ## Search Scope Search operates within the context of your current project: * Only documents in the active project are searched * All processed documents are included * Web sources added to the project are also searchable * Documents must be in **Ready** status to appear in search results Documents that are still processing will not appear in search results. Wait for the processing to complete before searching.