> ## Documentation Index > Fetch the complete documentation index at: https://docs.moorcheh.ai/llms.txt > Use this file to discover all available pages before exploring further. # Use Cases & Examples > Real-world examples and use cases for Moorcheh ## Overview Explore practical use cases and implementation examples for Moorcheh's semantic search and AI generation capabilities. ## RAG (Retrieval-Augmented Generation) Build intelligent Q\&A systems that combine semantic search with AI generation. ```python theme={null} from moorcheh_sdk import MoorchehClient client = MoorchehClient(api_key="your-api-key") # 1. Upload your knowledge base documents = [ { "id": "doc-1", "text": "Moorcheh uses MIB technology for fast semantic search.", "metadata": {"category": "technology", "source": "docs"} }, { "id": "doc-2", "text": "The API supports 9 AI models including Claude and GPT.", "metadata": {"category": "features", "source": "docs"} } ] client.upload_text(namespace_name="knowledge-base", documents=documents) # 2. Generate contextual answers answer = client.get_answer( namespace="knowledge-base", query="What technology does Moorcheh use?", ai_model="anthropic.claude-sonnet-4-6", top_k=3 ) print(answer["answer"]) ``` **Best for:** * Customer support chatbots * Internal knowledge bases * Documentation Q\&A * Educational platforms ## Document Search Create powerful search engines for large document collections. ```python theme={null} # Upload documents with metadata client.upload_text( namespace_name="legal-docs", documents=[ { "id": "contract-001", "text": "Service Agreement between Party A and Party B...", "metadata": { "doc_type": "contract", "year": "2024", "department": "legal" } } ] ) # Search with metadata filters results = client.search( namespaces=["legal-docs"], query="service level agreements #doc_type:contract #year:2024", top_k=10 ) for match in results["matches"]: print(f"Document: {match['id']}") print(f"Relevance: {match['score']}") print(f"Type: {match['metadata']['doc_type']}") ``` **Best for:** * Legal document search * Research paper databases * Corporate knowledge management * Content management systems ## Semantic Deduplication Identify and remove duplicate or similar content. ```python theme={null} # Upload content client.upload_text( namespace_name="articles", documents=[ {"id": "article-1", "text": "Introduction to machine learning..."}, {"id": "article-2", "text": "Machine learning basics..."}, {"id": "article-3", "text": "Getting started with ML..."} ] ) # Find similar articles def find_duplicates(article_id, threshold=0.85): # Get the article article = client.get_documents( namespace_name="articles", ids=[article_id] ) # Search for similar content results = client.search( namespaces=["articles"], query=article[0]["text"], top_k=10 ) # Filter by similarity threshold duplicates = [ match for match in results["matches"] if match["score"] > threshold and match["id"] != article_id ] return duplicates # Find duplicates duplicates = find_duplicates("article-1", threshold=0.85) print(f"Found {len(duplicates)} similar articles") ``` **Best for:** * Content moderation * Data cleaning * Duplicate detection * Content consolidation ## Recommendation System Build content or product recommendation engines. ```python theme={null} # Upload products with embeddings client.upload_vectors( namespace_name="products", vectors=[ { "id": "prod-001", "vector": [0.1, 0.2, ...], # Product embedding "name": "Wireless Headphones", "category": "electronics", "price": 99.99 } ] ) # Get recommendations based on user interest def get_recommendations(user_interest_vector, num_recommendations=5): results = client.search( namespaces=["products"], query=user_interest_vector, top_k=num_recommendations ) return [ { "name": match["name"], "score": match["score"], "category": match["category"] } for match in results["matches"] ] # Get personalized recommendations recommendations = get_recommendations(user_vector, num_recommendations=5) ``` **Best for:** * E-commerce recommendations * Content recommendations * Similar item suggestions * Personalization engines ## Multi-lingual Search Search across documents in multiple languages. ```python theme={null} # Upload documents in different languages documents = [ {"id": "doc-en", "text": "Welcome to Moorcheh", "metadata": {"lang": "en"}}, {"id": "doc-es", "text": "Bienvenido a Moorcheh", "metadata": {"lang": "es"}}, {"id": "doc-fr", "text": "Bienvenue à Moorcheh", "metadata": {"lang": "fr"}} ] client.upload_text(namespace_name="multi-lang-docs", documents=documents) # Search in specific language results = client.search( namespaces=["multi-lang-docs"], query="welcome #lang:en", top_k=5 ) # Or search across all languages results = client.search( namespaces=["multi-lang-docs"], query="welcome", top_k=5 ) ``` **Best for:** * International platforms * Multi-lingual support * Content localization * Global knowledge bases ## Code Search Search through code repositories and documentation. ```python theme={null} # Upload code snippets code_snippets = [ { "id": "snippet-1", "text": "def authenticate(api_key): return validate_key(api_key)", "metadata": {"language": "python", "function": "authentication"} }, { "id": "snippet-2", "text": "function search(query) { return api.search(query); }", "metadata": {"language": "javascript", "function": "search"} } ] client.upload_text(namespace_name="code-base", documents=code_snippets) # Search for code results = client.search( namespaces=["code-base"], query="how to authenticate users #language:python", top_k=5 ) # Generate code explanation answer = client.get_answer( namespace="code-base", query="Explain how authentication works in the codebase", ai_model="deepseek.r1-v1:0" # Best for code ) ``` **Best for:** * Code search engines * Developer documentation * Code examples search * Technical support ## Next Steps Get started with your first implementation Learn optimization techniques Explore all API endpoints Use with LangChain and LlamaIndex