Struct VectorStoreSearchBuilder 

pub struct VectorStoreSearchBuilder { /* private fields */ }

Builder for searching through vector stores.

Implementations

impl VectorStoreSearchBuilder

pub fn new(vector_store_id: impl Into<String>, query: impl Into<String>) -> Self

Create a new vector store search builder.

Examples found in repository

examples/vector_stores.rs (line 493)

fn run_advanced_search_patterns_example() -> Result<(), Error> {
    println!("\n Example 6: Advanced Search Patterns and Optimization");
    println!("{}", "=".repeat(60));

    // Create optimized search store
    let optimized_store = VectorStoreBuilder::new()
        .name("Advanced Search Optimization Store")
        .add_file("file-technical-docs-001")
        .add_file("file-user-feedback-002")
        .add_file("file-performance-data-003")
        .add_file("file-best-practices-004")
        .metadata("search_optimized", "true")
        .metadata("indexing", "enhanced")
        .metadata("caching", "enabled");

    println!(" Created advanced search store:");
    println!("   Optimization: Enhanced indexing");
    println!("   Caching: Enabled");
    println!("   Files: {} documents", optimized_store.file_count());

    // Demonstrate advanced search patterns
    println!("\n Advanced Search Patterns:");

    // Multi-stage search
    println!("   1.  Multi-stage Search:");
    println!("      Stage 1: Broad semantic search (100 results)");
    println!("      Stage 2: Filtered refinement (20 results)");
    println!("      Stage 3: Relevance re-ranking (5 top results)");

    let multi_stage_search =
        VectorStoreSearchBuilder::new("advanced-store-789", "machine learning best practices")
            .limit(100)
            .filter("category", "best_practices")
            .filter("verified", "true");

    println!("      Query: '{}'", multi_stage_search.query());
    println!(
        "      Initial limit: {}",
        multi_stage_search.limit_ref().unwrap()
    );

    // Contextual search
    println!("   2.  Contextual Search:");
    println!("      Context: User role, project phase, domain expertise");
    println!("      Adaptation: Results tailored to user context");

    let _contextual_search =
        search_vector_store_with_limit("advanced-store-789", "deployment strategies", 15)
            .filter("audience", "senior_engineer")
            .filter("complexity", "advanced")
            .filter("domain", "cloud_infrastructure");

    println!("      Audience: senior_engineer");
    println!("      Complexity: advanced");
    println!("      Domain: cloud_infrastructure");

    // Hybrid search approaches
    println!("   3.  Hybrid Search Approaches:");
    println!("      Semantic similarity + keyword matching");
    println!("      Vector search + traditional full-text search");
    println!("      AI-enhanced query understanding");

    // Search performance optimization
    println!("\n Search Performance Optimization:");
    println!("    Query optimization and caching");
    println!("    Result pre-computation for common queries");
    println!("    Incremental index updates");
    println!("    Load balancing across vector stores");
    println!("    Machine learning-based relevance tuning");

    // Quality metrics and monitoring
    println!("\n Search Quality Metrics:");
    println!("    Relevance scores and user feedback");
    println!("   ⏱ Query response time analysis");
    println!("    Search success rate tracking");
    println!("    Usage pattern analysis");
    println!("    Continuous improvement recommendations");

    Ok(())
}

pub fn limit(self, limit: i32) -> Self

Set the maximum number of results to return.

Examples found in repository

examples/vector_stores.rs (line 494)

fn run_advanced_search_patterns_example() -> Result<(), Error> {
    println!("\n Example 6: Advanced Search Patterns and Optimization");
    println!("{}", "=".repeat(60));

    // Create optimized search store
    let optimized_store = VectorStoreBuilder::new()
        .name("Advanced Search Optimization Store")
        .add_file("file-technical-docs-001")
        .add_file("file-user-feedback-002")
        .add_file("file-performance-data-003")
        .add_file("file-best-practices-004")
        .metadata("search_optimized", "true")
        .metadata("indexing", "enhanced")
        .metadata("caching", "enabled");

    println!(" Created advanced search store:");
    println!("   Optimization: Enhanced indexing");
    println!("   Caching: Enabled");
    println!("   Files: {} documents", optimized_store.file_count());

    // Demonstrate advanced search patterns
    println!("\n Advanced Search Patterns:");

    // Multi-stage search
    println!("   1.  Multi-stage Search:");
    println!("      Stage 1: Broad semantic search (100 results)");
    println!("      Stage 2: Filtered refinement (20 results)");
    println!("      Stage 3: Relevance re-ranking (5 top results)");

    let multi_stage_search =
        VectorStoreSearchBuilder::new("advanced-store-789", "machine learning best practices")
            .limit(100)
            .filter("category", "best_practices")
            .filter("verified", "true");

    println!("      Query: '{}'", multi_stage_search.query());
    println!(
        "      Initial limit: {}",
        multi_stage_search.limit_ref().unwrap()
    );

    // Contextual search
    println!("   2.  Contextual Search:");
    println!("      Context: User role, project phase, domain expertise");
    println!("      Adaptation: Results tailored to user context");

    let _contextual_search =
        search_vector_store_with_limit("advanced-store-789", "deployment strategies", 15)
            .filter("audience", "senior_engineer")
            .filter("complexity", "advanced")
            .filter("domain", "cloud_infrastructure");

    println!("      Audience: senior_engineer");
    println!("      Complexity: advanced");
    println!("      Domain: cloud_infrastructure");

    // Hybrid search approaches
    println!("   3.  Hybrid Search Approaches:");
    println!("      Semantic similarity + keyword matching");
    println!("      Vector search + traditional full-text search");
    println!("      AI-enhanced query understanding");

    // Search performance optimization
    println!("\n Search Performance Optimization:");
    println!("    Query optimization and caching");
    println!("    Result pre-computation for common queries");
    println!("    Incremental index updates");
    println!("    Load balancing across vector stores");
    println!("    Machine learning-based relevance tuning");

    // Quality metrics and monitoring
    println!("\n Search Quality Metrics:");
    println!("    Relevance scores and user feedback");
    println!("   ⏱ Query response time analysis");
    println!("    Search success rate tracking");
    println!("    Usage pattern analysis");
    println!("    Continuous improvement recommendations");

    Ok(())
}

pub fn filter(self, key: impl Into<String>, value: impl Into<String>) -> Self

Add a filter to the search.

Examples found in repository

examples/vector_stores.rs (line 268)

fn run_semantic_search_example() -> Result<(), Error> {
    println!("\n Example 3: Semantic Search and Similarity Queries");
    println!("{}", "=".repeat(60));

    // Create a search-optimized vector store
    let search_store = simple_vector_store("Semantic Search Demo Store")
        .add_file("file-ml-concepts-001")
        .add_file("file-nlp-techniques-002")
        .add_file("file-deep-learning-003")
        .add_file("file-computer-vision-004")
        .add_file("file-ai-ethics-005")
        .metadata("domain", "machine_learning")
        .metadata("search_optimized", "true");

    println!(" Created search-optimized vector store:");
    println!("   Name: {}", search_store.name_ref().unwrap());
    println!("   Domain: Machine Learning");
    println!("   Documents: {} files", search_store.file_count());

    // Demonstrate various search patterns
    println!("\n Search Query Examples:");

    // Basic semantic search
    let basic_search = search_vector_store("search-store-123", "neural network architectures");
    println!("   1. Basic Search:");
    println!("      Query: '{}'", basic_search.query());
    println!("      Store: {}", basic_search.vector_store_id());

    // Limited result search
    let limited_search = search_vector_store_with_limit(
        "search-store-123",
        "natural language processing techniques",
        5,
    );
    println!("   2. Limited Results:");
    println!("      Query: '{}'", limited_search.query());
    println!(
        "      Limit: {} results",
        limited_search.limit_ref().unwrap()
    );

    // Advanced filtered search
    let filtered_search =
        search_vector_store_with_limit("search-store-123", "computer vision applications", 10)
            .filter("category", "practical_applications")
            .filter("difficulty", "intermediate");

    println!("   3. Filtered Search:");
    println!("      Query: '{}'", filtered_search.query());
    println!(
        "      Filters: {} applied",
        filtered_search.filter_ref().len()
    );
    for (key, value) in filtered_search.filter_ref() {
        println!("         {}={}", key, value);
    }

    // Demonstrate search result processing
    println!("\n Search Result Processing:");
    println!("    Semantic similarity ranking");
    println!("    Document excerpt extraction");
    println!("    Relevance score calculation");
    println!("    Source location identification");
    println!("    Related content suggestions");

    // Show different query types
    println!("\n Query Type Examples:");
    let query_examples = vec![
        (
            "Conceptual",
            "What is machine learning?",
            "Broad conceptual understanding",
        ),
        (
            "Technical",
            "How to implement backpropagation?",
            "Specific technical implementation",
        ),
        (
            "Comparative",
            "LSTM vs Transformer architectures",
            "Comparative analysis",
        ),
        (
            "Problem-solving",
            "Overfitting in neural networks",
            "Problem identification and solutions",
        ),
        (
            "Application",
            "Computer vision in healthcare",
            "Domain-specific applications",
        ),
    ];

    for (query_type, query, description) in query_examples {
        println!("    {}: '{}'", query_type, query);
        println!("      Purpose: {}", description);
    }

    Ok(())
}

/// Example 4: Enterprise Knowledge Base
fn run_enterprise_knowledge_base_example() -> Result<(), Error> {
    println!("\n Example 4: Enterprise Knowledge Base");
    println!("{}", "=".repeat(60));

    // Create enterprise-scale vector stores
    let enterprise_stores = create_enterprise_knowledge_base()?;

    println!(" Enterprise Knowledge Base Architecture:");
    for (department, store) in enterprise_stores {
        println!("    {}", department);
        println!("      Files: {} documents", store.file_count());
        println!(
            "      Retention: {} days",
            store
                .expires_after_ref()
                .map_or("permanent".to_string(), |exp| exp.days.to_string())
                .as_str()
        );

        // Show metadata structure
        for (key, value) in store.metadata_ref() {
            println!("      {}: {}", key, value);
        }
        println!();
    }

    // Demonstrate cross-departmental search
    println!(" Cross-Departmental Search Examples:");

    let cross_searches = vec![
        (
            "Security Compliance",
            "GDPR data handling procedures",
            vec!["legal", "engineering", "hr"],
        ),
        (
            "Product Launch",
            "Q4 release planning and coordination",
            vec!["product", "engineering", "marketing"],
        ),
        (
            "Budget Planning",
            "Annual technology investment strategy",
            vec!["finance", "engineering", "executive"],
        ),
        (
            "Process Improvement",
            "Remote work productivity guidelines",
            vec!["hr", "operations", "it"],
        ),
    ];

    for (topic, query, departments) in cross_searches {
        println!("    {}: '{}'", topic, query);
        println!("      Search scope: {}", departments.join(", "));
    }

    println!("\n Enterprise Features:");
    println!("    Role-based access control");
    println!("    Usage analytics and monitoring");
    println!("    Automated content lifecycle management");
    println!("    Search performance optimization");
    println!("    Backup and disaster recovery");
    println!("    Compliance and audit trails");

    Ok(())
}

/// Example 5: Vector Store Lifecycle Management
fn run_vector_store_lifecycle_example() -> Result<(), Error> {
    println!("\n Example 5: Vector Store Lifecycle Management");
    println!("{}", "=".repeat(60));

    // Demonstrate different lifecycle patterns
    println!("⏰ Vector Store Lifecycle Patterns:");

    // Temporary stores for sessions
    let session_store = temporary_vector_store("User Session Store", 1)
        .add_file("file-session-context-001")
        .metadata("session_id", "sess_12345")
        .metadata("user_id", "user_67890");

    println!("    Session-based (1 day):");
    println!("      Purpose: Temporary user context");
    println!("      Files: {}", session_store.file_count());
    println!("      Auto-cleanup: ");

    // Project stores
    let project_store = temporary_vector_store("Project Alpha Documentation", 90)
        .add_file("file-project-spec-001")
        .add_file("file-meeting-notes-002")
        .add_file("file-progress-reports-003")
        .metadata("project_id", "proj_alpha_2024")
        .metadata("phase", "development");

    println!("    Project-based (90 days):");
    println!("      Purpose: Project lifecycle documentation");
    println!("      Files: {}", project_store.file_count());
    println!("      Cleanup: After project completion");

    // Long-term knowledge stores
    let knowledge_store = simple_vector_store("Institutional Knowledge Base")
        .add_file("file-company-history-001")
        .add_file("file-best-practices-002")
        .add_file("file-lessons-learned-003")
        .metadata("retention", "permanent")
        .metadata("backup", "enabled")
        .metadata("compliance", "required");

    println!("    Institutional (permanent):");
    println!("      Purpose: Long-term organizational knowledge");
    println!("      Files: {}", knowledge_store.file_count());
    println!("      Cleanup: Manual review only");

    // Demonstrate lifecycle events
    println!("\n Lifecycle Event Handling:");
    println!("    Creation: Automatic indexing and optimization");
    println!("    Updates: Incremental re-indexing of modified files");
    println!("    Monitoring: Usage tracking and performance metrics");
    println!("    Warnings: Expiration notifications and alerts");
    println!("    Cleanup: Automatic or manual deletion processes");
    println!("    Archival: Long-term storage for compliance");

    // Show cost optimization strategies
    println!("\n Cost Optimization Strategies:");
    println!("    Smart expiration policies based on usage");
    println!("    Analytics-driven storage optimization");
    println!("    Automatic compression for archived content");
    println!("    Tiered storage (hot, warm, cold)");
    println!("    Usage-based scaling recommendations");

    Ok(())
}

/// Example 6: Advanced Search Patterns and Optimization
fn run_advanced_search_patterns_example() -> Result<(), Error> {
    println!("\n Example 6: Advanced Search Patterns and Optimization");
    println!("{}", "=".repeat(60));

    // Create optimized search store
    let optimized_store = VectorStoreBuilder::new()
        .name("Advanced Search Optimization Store")
        .add_file("file-technical-docs-001")
        .add_file("file-user-feedback-002")
        .add_file("file-performance-data-003")
        .add_file("file-best-practices-004")
        .metadata("search_optimized", "true")
        .metadata("indexing", "enhanced")
        .metadata("caching", "enabled");

    println!(" Created advanced search store:");
    println!("   Optimization: Enhanced indexing");
    println!("   Caching: Enabled");
    println!("   Files: {} documents", optimized_store.file_count());

    // Demonstrate advanced search patterns
    println!("\n Advanced Search Patterns:");

    // Multi-stage search
    println!("   1.  Multi-stage Search:");
    println!("      Stage 1: Broad semantic search (100 results)");
    println!("      Stage 2: Filtered refinement (20 results)");
    println!("      Stage 3: Relevance re-ranking (5 top results)");

    let multi_stage_search =
        VectorStoreSearchBuilder::new("advanced-store-789", "machine learning best practices")
            .limit(100)
            .filter("category", "best_practices")
            .filter("verified", "true");

    println!("      Query: '{}'", multi_stage_search.query());
    println!(
        "      Initial limit: {}",
        multi_stage_search.limit_ref().unwrap()
    );

    // Contextual search
    println!("   2.  Contextual Search:");
    println!("      Context: User role, project phase, domain expertise");
    println!("      Adaptation: Results tailored to user context");

    let _contextual_search =
        search_vector_store_with_limit("advanced-store-789", "deployment strategies", 15)
            .filter("audience", "senior_engineer")
            .filter("complexity", "advanced")
            .filter("domain", "cloud_infrastructure");

    println!("      Audience: senior_engineer");
    println!("      Complexity: advanced");
    println!("      Domain: cloud_infrastructure");

    // Hybrid search approaches
    println!("   3.  Hybrid Search Approaches:");
    println!("      Semantic similarity + keyword matching");
    println!("      Vector search + traditional full-text search");
    println!("      AI-enhanced query understanding");

    // Search performance optimization
    println!("\n Search Performance Optimization:");
    println!("    Query optimization and caching");
    println!("    Result pre-computation for common queries");
    println!("    Incremental index updates");
    println!("    Load balancing across vector stores");
    println!("    Machine learning-based relevance tuning");

    // Quality metrics and monitoring
    println!("\n Search Quality Metrics:");
    println!("    Relevance scores and user feedback");
    println!("   ⏱ Query response time analysis");
    println!("    Search success rate tracking");
    println!("    Usage pattern analysis");
    println!("    Continuous improvement recommendations");

    Ok(())
}

examples/assistants_file_search.rs (line 256)

fn run_research_assistant_example() -> Result<(), Error> {
    println!("\n Example 3: Research Assistant with Advanced RAG");
    println!("{}", "=".repeat(60));

    // Create research-focused assistant
    let _research_assistant = assistant_with_instructions(
        "gpt-4-1106-preview",
        "Research Assistant",
        "You are a research assistant specializing in literature review and analysis. Help researchers find relevant information, identify patterns across documents, synthesize findings, and suggest research directions. Always provide comprehensive citations and acknowledge research limitations."
    )
    .add_tool(tool_file_search());

    println!(" Created research assistant:");
    println!("   Focus: Literature review and cross-document analysis");

    // Create a comprehensive research vector store
    let _research_store = VectorStoreBuilder::new()
        .name("Research Literature Database")
        .add_file("file-paper-ai-ethics-001")
        .add_file("file-paper-ml-bias-002")
        .add_file("file-paper-fairness-003")
        .add_file("file-survey-responsible-ai-004")
        .add_file("file-whitepaper-governance-005")
        .metadata("domain", "AI Ethics")
        .metadata("papers_count", "50")
        .metadata("date_range", "2020-2024");

    println!("\n Research Literature Database:");
    println!("   Domain: AI Ethics and Responsible AI");
    println!("   Papers: 5 documents loaded (representing 50 papers)");
    println!("   Date range: 2020-2024");

    println!("\n Research Query:");
    println!(
        "   'What are the current approaches to addressing algorithmic bias in machine learning?'"
    );
    println!("   'Please provide a comprehensive overview with citations.'");

    println!("\n Advanced RAG Research Workflow:");
    println!("   1.  Query analysis and decomposition");
    println!("   2.  Multi-faceted search across all documents");
    println!("   3.  Semantic clustering of results");
    println!("   4.  Cross-reference analysis between papers");
    println!("   5.  Identify trends and patterns");
    println!("   6.  Synthesize comprehensive overview");
    println!("   7.  Provide detailed citations and references");

    // Demonstrate search refinement
    let refined_search =
        search_vector_store_with_limit("research-db-123", "algorithmic bias mitigation", 20)
            .filter("category", "methodology")
            .filter("confidence", "high");

    println!("\n Search Refinement:");
    println!("   Query: algorithmic bias mitigation");
    println!("   Limit: {} results", refined_search.limit_ref().unwrap());
    println!("   Filters: category=methodology, confidence=high");

    println!("\n Expected Research Response:");
    println!("    Executive Summary:");
    println!("      • Overview of current bias mitigation approaches");
    println!("      • Key methodological categories identified");
    println!("      • Emerging trends and best practices");
    println!("   ");
    println!("    Detailed Analysis:");
    println!("      • Pre-processing techniques (data augmentation, sampling)");
    println!("      • In-processing methods (fairness constraints, adversarial training)");
    println!("      • Post-processing approaches (threshold optimization, calibration)");
    println!("   ");
    println!("    Comprehensive Citations:");
    println!("      • [Smith et al., 2023] - Fairness constraints in ML training");
    println!("      • [Johnson & Lee, 2024] - Bias detection in neural networks");
    println!("      • [Research Survey, 2024] - Comprehensive bias mitigation review");
    println!("   ");
    println!("    Future Research Directions:");
    println!("      • Intersectional bias analysis");
    println!("      • Real-time bias monitoring");
    println!("      • Domain-specific mitigation strategies");

    Ok(())
}

/// Example 4: Citation and Source Attribution
fn run_citation_example() -> Result<(), Error> {
    println!("\n Example 4: Citation and Source Attribution");
    println!("{}", "=".repeat(60));

    // Create citation-focused assistant
    let _citation_assistant = AssistantBuilder::new("gpt-4-1106-preview")
        .name("Citation Specialist")
        .description("Provides detailed source attribution and citation management")
        .instructions(
            "You are a citation specialist. Always provide accurate, detailed citations for any information retrieved from documents. Use proper academic citation formats, include page numbers when available, and distinguish between direct quotes and paraphrased content."
        )
        .add_tool(tool_file_search());

    println!(" Created citation specialist assistant:");
    println!("   Focus: Accurate source attribution and citation formatting");

    // Create thread for citation-heavy work
    let _citation_thread = simple_thread()
        .metadata("citation_style", "APA")
        .metadata("requirement", "academic_standards")
        .metadata("verification", "enabled");

    println!("\n Citation Requirements:");
    println!("   Style: APA format");
    println!("   Standards: Academic-level accuracy");
    println!("   Verification: Enabled for all sources");

    println!("\n Citation Query:");
    println!("   'Provide a summary of the key arguments about privacy in AI systems,'");
    println!("   'with detailed citations for each point made.'");

    println!("\n Citation-Focused RAG Workflow:");
    println!("   1.  Search for relevant content across documents");
    println!("   2.  Extract content with precise location data");
    println!("   3.  Generate response with inline citations");
    println!("   4.  Verify citation accuracy and completeness");
    println!("   5.  Format citations according to specified style");
    println!("   6.  Cross-check for citation consistency");

    // Demonstrate different citation formats
    println!("\n Citation Format Examples:");
    println!("    Direct Quote:");
    println!(
        "      \"Privacy-preserving AI requires careful balance between utility and protection\""
    );
    println!("      (Johnson, 2024, p. 15)");
    println!("   ");
    println!("    Paraphrased Content:");
    println!("      Recent research indicates that differential privacy methods show promise");
    println!("      for protecting individual data in ML training (Smith & Lee, 2023).");
    println!("   ");
    println!("    Multiple Source Synthesis:");
    println!("      Several studies have demonstrated the effectiveness of federated learning");
    println!("      approaches (Chen et al., 2023; Rodriguez, 2024; Brown & Wilson, 2023).");

    println!("\n Expected Citation Response:");
    println!("    Structured Summary with Citations:");
    println!("      • Privacy challenges in AI systems (Anderson, 2024, pp. 23-25)");
    println!("      • Technical solutions: differential privacy (Johnson et al., 2023)");
    println!("      • Regulatory considerations (Privacy Commission Report, 2024, §3.2)");
    println!("   ");
    println!("    Reference List:");
    println!(
        "      Anderson, M. (2024). AI Privacy Challenges. Tech Ethics Journal, 15(3), 20-30."
    );
    println!("      Johnson, P., Smith, R., & Lee, K. (2023). Differential privacy in practice.");
    println!("      Privacy Commission. (2024). AI governance guidelines (Report #2024-AI-001).");

    Ok(())
}

/// Example 5: Multi-Document Analysis and Synthesis
fn run_multi_document_analysis_example() -> Result<(), Error> {
    println!("\n Example 5: Multi-Document Analysis and Synthesis");
    println!("{}", "=".repeat(60));

    // Create multi-document analysis assistant
    let _analysis_assistant = assistant_with_instructions(
        "gpt-4-1106-preview",
        "Document Analysis Specialist",
        "You are a document analysis expert. Compare and contrast information across multiple documents, identify contradictions or gaps, synthesize information from diverse sources, and provide comprehensive analysis that considers multiple perspectives."
    )
    .add_tool(tool_file_search());

    println!(" Created document analysis assistant:");
    println!("   Specialty: Cross-document comparison and synthesis");

    // Create comprehensive document store for analysis
    let _analysis_store = VectorStoreBuilder::new()
        .name("Multi-Document Analysis Store")
        .add_file("file-policy-proposal-v1")
        .add_file("file-policy-proposal-v2")
        .add_file("file-stakeholder-feedback-001")
        .add_file("file-legal-review-002")
        .add_file("file-technical-assessment-003")
        .add_file("file-cost-benefit-analysis-004")
        .metadata("analysis_type", "policy_comparison")
        .metadata("documents", "6")
        .metadata("perspectives", "multiple");

    println!("\n Multi-Document Analysis Setup:");
    println!("   Documents: 6 files representing different perspectives");
    println!("   Analysis type: Policy proposal comparison");
    println!("   Perspectives: Technical, legal, financial, stakeholder");

    println!("\n Multi-Document Analysis Query:");
    println!("   'Compare the two policy proposals and analyze how stakeholder feedback'");
    println!("   'has been incorporated. Identify any conflicts between the legal review'");
    println!("   'and technical assessment.'");

    println!("\n Multi-Document RAG Analysis Workflow:");
    println!("   1.  Identify key comparison dimensions");
    println!("   2.  Search each document type systematically");
    println!("   3.  Create comparison matrix across documents");
    println!("   4.  Identify conflicts and contradictions");
    println!("   5.  Find connections and dependencies");
    println!("   6.  Synthesize comprehensive analysis");
    println!("   7.  Provide recommendations based on synthesis");

    // Demonstrate advanced search patterns
    let _comparative_search =
        search_vector_store_with_limit("analysis-store-456", "risk assessment comparison", 15)
            .filter("document_type", "technical,legal")
            .filter("section", "risks");

    println!("\n Advanced Search Pattern:");
    println!("   Query: risk assessment comparison");
    println!("   Target documents: technical + legal reviews");
    println!("   Focus section: risk analysis sections");

    println!("\n Expected Multi-Document Analysis:");
    println!("    Comparative Analysis:");
    println!("      Policy Proposal Comparison:");
    println!("      • V1 focuses on immediate implementation (Technical Assessment)");
    println!("      • V2 incorporates phased approach (Stakeholder Feedback)");
    println!("   ");
    println!("    Identified Conflicts:");
    println!("      • Legal review flags compliance issues with V1 approach");
    println!("      • Technical assessment questions feasibility of V2 timeline");
    println!("      • Cost analysis shows budget misalignment between proposals");
    println!("   ");
    println!("    Stakeholder Integration:");
    println!("      • 73% of feedback incorporated in V2 (Stakeholder Feedback doc)");
    println!("      • Privacy concerns addressed through technical modifications");
    println!("      • Cost concerns partially resolved via phased implementation");
    println!("   ");
    println!("    Synthesis Recommendations:");
    println!("      • Hybrid approach combining V1 technical framework with V2 timeline");
    println!("      • Address legal compliance through additional technical review");
    println!("      • Require budget revision to align with stakeholder expectations");

    Ok(())
}

pub fn vector_store_id(&self) -> &str

Get the vector store ID for this search.

Examples found in repository

examples/vector_stores.rs (line 250)

fn run_semantic_search_example() -> Result<(), Error> {
    println!("\n Example 3: Semantic Search and Similarity Queries");
    println!("{}", "=".repeat(60));

    // Create a search-optimized vector store
    let search_store = simple_vector_store("Semantic Search Demo Store")
        .add_file("file-ml-concepts-001")
        .add_file("file-nlp-techniques-002")
        .add_file("file-deep-learning-003")
        .add_file("file-computer-vision-004")
        .add_file("file-ai-ethics-005")
        .metadata("domain", "machine_learning")
        .metadata("search_optimized", "true");

    println!(" Created search-optimized vector store:");
    println!("   Name: {}", search_store.name_ref().unwrap());
    println!("   Domain: Machine Learning");
    println!("   Documents: {} files", search_store.file_count());

    // Demonstrate various search patterns
    println!("\n Search Query Examples:");

    // Basic semantic search
    let basic_search = search_vector_store("search-store-123", "neural network architectures");
    println!("   1. Basic Search:");
    println!("      Query: '{}'", basic_search.query());
    println!("      Store: {}", basic_search.vector_store_id());

    // Limited result search
    let limited_search = search_vector_store_with_limit(
        "search-store-123",
        "natural language processing techniques",
        5,
    );
    println!("   2. Limited Results:");
    println!("      Query: '{}'", limited_search.query());
    println!(
        "      Limit: {} results",
        limited_search.limit_ref().unwrap()
    );

    // Advanced filtered search
    let filtered_search =
        search_vector_store_with_limit("search-store-123", "computer vision applications", 10)
            .filter("category", "practical_applications")
            .filter("difficulty", "intermediate");

    println!("   3. Filtered Search:");
    println!("      Query: '{}'", filtered_search.query());
    println!(
        "      Filters: {} applied",
        filtered_search.filter_ref().len()
    );
    for (key, value) in filtered_search.filter_ref() {
        println!("         {}={}", key, value);
    }

    // Demonstrate search result processing
    println!("\n Search Result Processing:");
    println!("    Semantic similarity ranking");
    println!("    Document excerpt extraction");
    println!("    Relevance score calculation");
    println!("    Source location identification");
    println!("    Related content suggestions");

    // Show different query types
    println!("\n Query Type Examples:");
    let query_examples = vec![
        (
            "Conceptual",
            "What is machine learning?",
            "Broad conceptual understanding",
        ),
        (
            "Technical",
            "How to implement backpropagation?",
            "Specific technical implementation",
        ),
        (
            "Comparative",
            "LSTM vs Transformer architectures",
            "Comparative analysis",
        ),
        (
            "Problem-solving",
            "Overfitting in neural networks",
            "Problem identification and solutions",
        ),
        (
            "Application",
            "Computer vision in healthcare",
            "Domain-specific applications",
        ),
    ];

    for (query_type, query, description) in query_examples {
        println!("    {}: '{}'", query_type, query);
        println!("      Purpose: {}", description);
    }

    Ok(())
}

pub fn query(&self) -> &str

Get the search query.

Examples found in repository

examples/vector_stores.rs (line 249)

fn run_semantic_search_example() -> Result<(), Error> {
    println!("\n Example 3: Semantic Search and Similarity Queries");
    println!("{}", "=".repeat(60));

    // Create a search-optimized vector store
    let search_store = simple_vector_store("Semantic Search Demo Store")
        .add_file("file-ml-concepts-001")
        .add_file("file-nlp-techniques-002")
        .add_file("file-deep-learning-003")
        .add_file("file-computer-vision-004")
        .add_file("file-ai-ethics-005")
        .metadata("domain", "machine_learning")
        .metadata("search_optimized", "true");

    println!(" Created search-optimized vector store:");
    println!("   Name: {}", search_store.name_ref().unwrap());
    println!("   Domain: Machine Learning");
    println!("   Documents: {} files", search_store.file_count());

    // Demonstrate various search patterns
    println!("\n Search Query Examples:");

    // Basic semantic search
    let basic_search = search_vector_store("search-store-123", "neural network architectures");
    println!("   1. Basic Search:");
    println!("      Query: '{}'", basic_search.query());
    println!("      Store: {}", basic_search.vector_store_id());

    // Limited result search
    let limited_search = search_vector_store_with_limit(
        "search-store-123",
        "natural language processing techniques",
        5,
    );
    println!("   2. Limited Results:");
    println!("      Query: '{}'", limited_search.query());
    println!(
        "      Limit: {} results",
        limited_search.limit_ref().unwrap()
    );

    // Advanced filtered search
    let filtered_search =
        search_vector_store_with_limit("search-store-123", "computer vision applications", 10)
            .filter("category", "practical_applications")
            .filter("difficulty", "intermediate");

    println!("   3. Filtered Search:");
    println!("      Query: '{}'", filtered_search.query());
    println!(
        "      Filters: {} applied",
        filtered_search.filter_ref().len()
    );
    for (key, value) in filtered_search.filter_ref() {
        println!("         {}={}", key, value);
    }

    // Demonstrate search result processing
    println!("\n Search Result Processing:");
    println!("    Semantic similarity ranking");
    println!("    Document excerpt extraction");
    println!("    Relevance score calculation");
    println!("    Source location identification");
    println!("    Related content suggestions");

    // Show different query types
    println!("\n Query Type Examples:");
    let query_examples = vec![
        (
            "Conceptual",
            "What is machine learning?",
            "Broad conceptual understanding",
        ),
        (
            "Technical",
            "How to implement backpropagation?",
            "Specific technical implementation",
        ),
        (
            "Comparative",
            "LSTM vs Transformer architectures",
            "Comparative analysis",
        ),
        (
            "Problem-solving",
            "Overfitting in neural networks",
            "Problem identification and solutions",
        ),
        (
            "Application",
            "Computer vision in healthcare",
            "Domain-specific applications",
        ),
    ];

    for (query_type, query, description) in query_examples {
        println!("    {}: '{}'", query_type, query);
        println!("      Purpose: {}", description);
    }

    Ok(())
}

/// Example 4: Enterprise Knowledge Base
fn run_enterprise_knowledge_base_example() -> Result<(), Error> {
    println!("\n Example 4: Enterprise Knowledge Base");
    println!("{}", "=".repeat(60));

    // Create enterprise-scale vector stores
    let enterprise_stores = create_enterprise_knowledge_base()?;

    println!(" Enterprise Knowledge Base Architecture:");
    for (department, store) in enterprise_stores {
        println!("    {}", department);
        println!("      Files: {} documents", store.file_count());
        println!(
            "      Retention: {} days",
            store
                .expires_after_ref()
                .map_or("permanent".to_string(), |exp| exp.days.to_string())
                .as_str()
        );

        // Show metadata structure
        for (key, value) in store.metadata_ref() {
            println!("      {}: {}", key, value);
        }
        println!();
    }

    // Demonstrate cross-departmental search
    println!(" Cross-Departmental Search Examples:");

    let cross_searches = vec![
        (
            "Security Compliance",
            "GDPR data handling procedures",
            vec!["legal", "engineering", "hr"],
        ),
        (
            "Product Launch",
            "Q4 release planning and coordination",
            vec!["product", "engineering", "marketing"],
        ),
        (
            "Budget Planning",
            "Annual technology investment strategy",
            vec!["finance", "engineering", "executive"],
        ),
        (
            "Process Improvement",
            "Remote work productivity guidelines",
            vec!["hr", "operations", "it"],
        ),
    ];

    for (topic, query, departments) in cross_searches {
        println!("    {}: '{}'", topic, query);
        println!("      Search scope: {}", departments.join(", "));
    }

    println!("\n Enterprise Features:");
    println!("    Role-based access control");
    println!("    Usage analytics and monitoring");
    println!("    Automated content lifecycle management");
    println!("    Search performance optimization");
    println!("    Backup and disaster recovery");
    println!("    Compliance and audit trails");

    Ok(())
}

/// Example 5: Vector Store Lifecycle Management
fn run_vector_store_lifecycle_example() -> Result<(), Error> {
    println!("\n Example 5: Vector Store Lifecycle Management");
    println!("{}", "=".repeat(60));

    // Demonstrate different lifecycle patterns
    println!("⏰ Vector Store Lifecycle Patterns:");

    // Temporary stores for sessions
    let session_store = temporary_vector_store("User Session Store", 1)
        .add_file("file-session-context-001")
        .metadata("session_id", "sess_12345")
        .metadata("user_id", "user_67890");

    println!("    Session-based (1 day):");
    println!("      Purpose: Temporary user context");
    println!("      Files: {}", session_store.file_count());
    println!("      Auto-cleanup: ");

    // Project stores
    let project_store = temporary_vector_store("Project Alpha Documentation", 90)
        .add_file("file-project-spec-001")
        .add_file("file-meeting-notes-002")
        .add_file("file-progress-reports-003")
        .metadata("project_id", "proj_alpha_2024")
        .metadata("phase", "development");

    println!("    Project-based (90 days):");
    println!("      Purpose: Project lifecycle documentation");
    println!("      Files: {}", project_store.file_count());
    println!("      Cleanup: After project completion");

    // Long-term knowledge stores
    let knowledge_store = simple_vector_store("Institutional Knowledge Base")
        .add_file("file-company-history-001")
        .add_file("file-best-practices-002")
        .add_file("file-lessons-learned-003")
        .metadata("retention", "permanent")
        .metadata("backup", "enabled")
        .metadata("compliance", "required");

    println!("    Institutional (permanent):");
    println!("      Purpose: Long-term organizational knowledge");
    println!("      Files: {}", knowledge_store.file_count());
    println!("      Cleanup: Manual review only");

    // Demonstrate lifecycle events
    println!("\n Lifecycle Event Handling:");
    println!("    Creation: Automatic indexing and optimization");
    println!("    Updates: Incremental re-indexing of modified files");
    println!("    Monitoring: Usage tracking and performance metrics");
    println!("    Warnings: Expiration notifications and alerts");
    println!("    Cleanup: Automatic or manual deletion processes");
    println!("    Archival: Long-term storage for compliance");

    // Show cost optimization strategies
    println!("\n Cost Optimization Strategies:");
    println!("    Smart expiration policies based on usage");
    println!("    Analytics-driven storage optimization");
    println!("    Automatic compression for archived content");
    println!("    Tiered storage (hot, warm, cold)");
    println!("    Usage-based scaling recommendations");

    Ok(())
}

/// Example 6: Advanced Search Patterns and Optimization
fn run_advanced_search_patterns_example() -> Result<(), Error> {
    println!("\n Example 6: Advanced Search Patterns and Optimization");
    println!("{}", "=".repeat(60));

    // Create optimized search store
    let optimized_store = VectorStoreBuilder::new()
        .name("Advanced Search Optimization Store")
        .add_file("file-technical-docs-001")
        .add_file("file-user-feedback-002")
        .add_file("file-performance-data-003")
        .add_file("file-best-practices-004")
        .metadata("search_optimized", "true")
        .metadata("indexing", "enhanced")
        .metadata("caching", "enabled");

    println!(" Created advanced search store:");
    println!("   Optimization: Enhanced indexing");
    println!("   Caching: Enabled");
    println!("   Files: {} documents", optimized_store.file_count());

    // Demonstrate advanced search patterns
    println!("\n Advanced Search Patterns:");

    // Multi-stage search
    println!("   1.  Multi-stage Search:");
    println!("      Stage 1: Broad semantic search (100 results)");
    println!("      Stage 2: Filtered refinement (20 results)");
    println!("      Stage 3: Relevance re-ranking (5 top results)");

    let multi_stage_search =
        VectorStoreSearchBuilder::new("advanced-store-789", "machine learning best practices")
            .limit(100)
            .filter("category", "best_practices")
            .filter("verified", "true");

    println!("      Query: '{}'", multi_stage_search.query());
    println!(
        "      Initial limit: {}",
        multi_stage_search.limit_ref().unwrap()
    );

    // Contextual search
    println!("   2.  Contextual Search:");
    println!("      Context: User role, project phase, domain expertise");
    println!("      Adaptation: Results tailored to user context");

    let _contextual_search =
        search_vector_store_with_limit("advanced-store-789", "deployment strategies", 15)
            .filter("audience", "senior_engineer")
            .filter("complexity", "advanced")
            .filter("domain", "cloud_infrastructure");

    println!("      Audience: senior_engineer");
    println!("      Complexity: advanced");
    println!("      Domain: cloud_infrastructure");

    // Hybrid search approaches
    println!("   3.  Hybrid Search Approaches:");
    println!("      Semantic similarity + keyword matching");
    println!("      Vector search + traditional full-text search");
    println!("      AI-enhanced query understanding");

    // Search performance optimization
    println!("\n Search Performance Optimization:");
    println!("    Query optimization and caching");
    println!("    Result pre-computation for common queries");
    println!("    Incremental index updates");
    println!("    Load balancing across vector stores");
    println!("    Machine learning-based relevance tuning");

    // Quality metrics and monitoring
    println!("\n Search Quality Metrics:");
    println!("    Relevance scores and user feedback");
    println!("   ⏱ Query response time analysis");
    println!("    Search success rate tracking");
    println!("    Usage pattern analysis");
    println!("    Continuous improvement recommendations");

    Ok(())
}

pub fn limit_ref(&self) -> Option<i32>

Get the search limit.

Examples found in repository

examples/vector_stores.rs (line 262)

fn run_semantic_search_example() -> Result<(), Error> {
    println!("\n Example 3: Semantic Search and Similarity Queries");
    println!("{}", "=".repeat(60));

    // Create a search-optimized vector store
    let search_store = simple_vector_store("Semantic Search Demo Store")
        .add_file("file-ml-concepts-001")
        .add_file("file-nlp-techniques-002")
        .add_file("file-deep-learning-003")
        .add_file("file-computer-vision-004")
        .add_file("file-ai-ethics-005")
        .metadata("domain", "machine_learning")
        .metadata("search_optimized", "true");

    println!(" Created search-optimized vector store:");
    println!("   Name: {}", search_store.name_ref().unwrap());
    println!("   Domain: Machine Learning");
    println!("   Documents: {} files", search_store.file_count());

    // Demonstrate various search patterns
    println!("\n Search Query Examples:");

    // Basic semantic search
    let basic_search = search_vector_store("search-store-123", "neural network architectures");
    println!("   1. Basic Search:");
    println!("      Query: '{}'", basic_search.query());
    println!("      Store: {}", basic_search.vector_store_id());

    // Limited result search
    let limited_search = search_vector_store_with_limit(
        "search-store-123",
        "natural language processing techniques",
        5,
    );
    println!("   2. Limited Results:");
    println!("      Query: '{}'", limited_search.query());
    println!(
        "      Limit: {} results",
        limited_search.limit_ref().unwrap()
    );

    // Advanced filtered search
    let filtered_search =
        search_vector_store_with_limit("search-store-123", "computer vision applications", 10)
            .filter("category", "practical_applications")
            .filter("difficulty", "intermediate");

    println!("   3. Filtered Search:");
    println!("      Query: '{}'", filtered_search.query());
    println!(
        "      Filters: {} applied",
        filtered_search.filter_ref().len()
    );
    for (key, value) in filtered_search.filter_ref() {
        println!("         {}={}", key, value);
    }

    // Demonstrate search result processing
    println!("\n Search Result Processing:");
    println!("    Semantic similarity ranking");
    println!("    Document excerpt extraction");
    println!("    Relevance score calculation");
    println!("    Source location identification");
    println!("    Related content suggestions");

    // Show different query types
    println!("\n Query Type Examples:");
    let query_examples = vec![
        (
            "Conceptual",
            "What is machine learning?",
            "Broad conceptual understanding",
        ),
        (
            "Technical",
            "How to implement backpropagation?",
            "Specific technical implementation",
        ),
        (
            "Comparative",
            "LSTM vs Transformer architectures",
            "Comparative analysis",
        ),
        (
            "Problem-solving",
            "Overfitting in neural networks",
            "Problem identification and solutions",
        ),
        (
            "Application",
            "Computer vision in healthcare",
            "Domain-specific applications",
        ),
    ];

    for (query_type, query, description) in query_examples {
        println!("    {}: '{}'", query_type, query);
        println!("      Purpose: {}", description);
    }

    Ok(())
}

/// Example 4: Enterprise Knowledge Base
fn run_enterprise_knowledge_base_example() -> Result<(), Error> {
    println!("\n Example 4: Enterprise Knowledge Base");
    println!("{}", "=".repeat(60));

    // Create enterprise-scale vector stores
    let enterprise_stores = create_enterprise_knowledge_base()?;

    println!(" Enterprise Knowledge Base Architecture:");
    for (department, store) in enterprise_stores {
        println!("    {}", department);
        println!("      Files: {} documents", store.file_count());
        println!(
            "      Retention: {} days",
            store
                .expires_after_ref()
                .map_or("permanent".to_string(), |exp| exp.days.to_string())
                .as_str()
        );

        // Show metadata structure
        for (key, value) in store.metadata_ref() {
            println!("      {}: {}", key, value);
        }
        println!();
    }

    // Demonstrate cross-departmental search
    println!(" Cross-Departmental Search Examples:");

    let cross_searches = vec![
        (
            "Security Compliance",
            "GDPR data handling procedures",
            vec!["legal", "engineering", "hr"],
        ),
        (
            "Product Launch",
            "Q4 release planning and coordination",
            vec!["product", "engineering", "marketing"],
        ),
        (
            "Budget Planning",
            "Annual technology investment strategy",
            vec!["finance", "engineering", "executive"],
        ),
        (
            "Process Improvement",
            "Remote work productivity guidelines",
            vec!["hr", "operations", "it"],
        ),
    ];

    for (topic, query, departments) in cross_searches {
        println!("    {}: '{}'", topic, query);
        println!("      Search scope: {}", departments.join(", "));
    }

    println!("\n Enterprise Features:");
    println!("    Role-based access control");
    println!("    Usage analytics and monitoring");
    println!("    Automated content lifecycle management");
    println!("    Search performance optimization");
    println!("    Backup and disaster recovery");
    println!("    Compliance and audit trails");

    Ok(())
}

/// Example 5: Vector Store Lifecycle Management
fn run_vector_store_lifecycle_example() -> Result<(), Error> {
    println!("\n Example 5: Vector Store Lifecycle Management");
    println!("{}", "=".repeat(60));

    // Demonstrate different lifecycle patterns
    println!("⏰ Vector Store Lifecycle Patterns:");

    // Temporary stores for sessions
    let session_store = temporary_vector_store("User Session Store", 1)
        .add_file("file-session-context-001")
        .metadata("session_id", "sess_12345")
        .metadata("user_id", "user_67890");

    println!("    Session-based (1 day):");
    println!("      Purpose: Temporary user context");
    println!("      Files: {}", session_store.file_count());
    println!("      Auto-cleanup: ");

    // Project stores
    let project_store = temporary_vector_store("Project Alpha Documentation", 90)
        .add_file("file-project-spec-001")
        .add_file("file-meeting-notes-002")
        .add_file("file-progress-reports-003")
        .metadata("project_id", "proj_alpha_2024")
        .metadata("phase", "development");

    println!("    Project-based (90 days):");
    println!("      Purpose: Project lifecycle documentation");
    println!("      Files: {}", project_store.file_count());
    println!("      Cleanup: After project completion");

    // Long-term knowledge stores
    let knowledge_store = simple_vector_store("Institutional Knowledge Base")
        .add_file("file-company-history-001")
        .add_file("file-best-practices-002")
        .add_file("file-lessons-learned-003")
        .metadata("retention", "permanent")
        .metadata("backup", "enabled")
        .metadata("compliance", "required");

    println!("    Institutional (permanent):");
    println!("      Purpose: Long-term organizational knowledge");
    println!("      Files: {}", knowledge_store.file_count());
    println!("      Cleanup: Manual review only");

    // Demonstrate lifecycle events
    println!("\n Lifecycle Event Handling:");
    println!("    Creation: Automatic indexing and optimization");
    println!("    Updates: Incremental re-indexing of modified files");
    println!("    Monitoring: Usage tracking and performance metrics");
    println!("    Warnings: Expiration notifications and alerts");
    println!("    Cleanup: Automatic or manual deletion processes");
    println!("    Archival: Long-term storage for compliance");

    // Show cost optimization strategies
    println!("\n Cost Optimization Strategies:");
    println!("    Smart expiration policies based on usage");
    println!("    Analytics-driven storage optimization");
    println!("    Automatic compression for archived content");
    println!("    Tiered storage (hot, warm, cold)");
    println!("    Usage-based scaling recommendations");

    Ok(())
}

/// Example 6: Advanced Search Patterns and Optimization
fn run_advanced_search_patterns_example() -> Result<(), Error> {
    println!("\n Example 6: Advanced Search Patterns and Optimization");
    println!("{}", "=".repeat(60));

    // Create optimized search store
    let optimized_store = VectorStoreBuilder::new()
        .name("Advanced Search Optimization Store")
        .add_file("file-technical-docs-001")
        .add_file("file-user-feedback-002")
        .add_file("file-performance-data-003")
        .add_file("file-best-practices-004")
        .metadata("search_optimized", "true")
        .metadata("indexing", "enhanced")
        .metadata("caching", "enabled");

    println!(" Created advanced search store:");
    println!("   Optimization: Enhanced indexing");
    println!("   Caching: Enabled");
    println!("   Files: {} documents", optimized_store.file_count());

    // Demonstrate advanced search patterns
    println!("\n Advanced Search Patterns:");

    // Multi-stage search
    println!("   1.  Multi-stage Search:");
    println!("      Stage 1: Broad semantic search (100 results)");
    println!("      Stage 2: Filtered refinement (20 results)");
    println!("      Stage 3: Relevance re-ranking (5 top results)");

    let multi_stage_search =
        VectorStoreSearchBuilder::new("advanced-store-789", "machine learning best practices")
            .limit(100)
            .filter("category", "best_practices")
            .filter("verified", "true");

    println!("      Query: '{}'", multi_stage_search.query());
    println!(
        "      Initial limit: {}",
        multi_stage_search.limit_ref().unwrap()
    );

    // Contextual search
    println!("   2.  Contextual Search:");
    println!("      Context: User role, project phase, domain expertise");
    println!("      Adaptation: Results tailored to user context");

    let _contextual_search =
        search_vector_store_with_limit("advanced-store-789", "deployment strategies", 15)
            .filter("audience", "senior_engineer")
            .filter("complexity", "advanced")
            .filter("domain", "cloud_infrastructure");

    println!("      Audience: senior_engineer");
    println!("      Complexity: advanced");
    println!("      Domain: cloud_infrastructure");

    // Hybrid search approaches
    println!("   3.  Hybrid Search Approaches:");
    println!("      Semantic similarity + keyword matching");
    println!("      Vector search + traditional full-text search");
    println!("      AI-enhanced query understanding");

    // Search performance optimization
    println!("\n Search Performance Optimization:");
    println!("    Query optimization and caching");
    println!("    Result pre-computation for common queries");
    println!("    Incremental index updates");
    println!("    Load balancing across vector stores");
    println!("    Machine learning-based relevance tuning");

    // Quality metrics and monitoring
    println!("\n Search Quality Metrics:");
    println!("    Relevance scores and user feedback");
    println!("   ⏱ Query response time analysis");
    println!("    Search success rate tracking");
    println!("    Usage pattern analysis");
    println!("    Continuous improvement recommendations");

    Ok(())
}

examples/assistants_file_search.rs (line 261)

fn run_research_assistant_example() -> Result<(), Error> {
    println!("\n Example 3: Research Assistant with Advanced RAG");
    println!("{}", "=".repeat(60));

    // Create research-focused assistant
    let _research_assistant = assistant_with_instructions(
        "gpt-4-1106-preview",
        "Research Assistant",
        "You are a research assistant specializing in literature review and analysis. Help researchers find relevant information, identify patterns across documents, synthesize findings, and suggest research directions. Always provide comprehensive citations and acknowledge research limitations."
    )
    .add_tool(tool_file_search());

    println!(" Created research assistant:");
    println!("   Focus: Literature review and cross-document analysis");

    // Create a comprehensive research vector store
    let _research_store = VectorStoreBuilder::new()
        .name("Research Literature Database")
        .add_file("file-paper-ai-ethics-001")
        .add_file("file-paper-ml-bias-002")
        .add_file("file-paper-fairness-003")
        .add_file("file-survey-responsible-ai-004")
        .add_file("file-whitepaper-governance-005")
        .metadata("domain", "AI Ethics")
        .metadata("papers_count", "50")
        .metadata("date_range", "2020-2024");

    println!("\n Research Literature Database:");
    println!("   Domain: AI Ethics and Responsible AI");
    println!("   Papers: 5 documents loaded (representing 50 papers)");
    println!("   Date range: 2020-2024");

    println!("\n Research Query:");
    println!(
        "   'What are the current approaches to addressing algorithmic bias in machine learning?'"
    );
    println!("   'Please provide a comprehensive overview with citations.'");

    println!("\n Advanced RAG Research Workflow:");
    println!("   1.  Query analysis and decomposition");
    println!("   2.  Multi-faceted search across all documents");
    println!("   3.  Semantic clustering of results");
    println!("   4.  Cross-reference analysis between papers");
    println!("   5.  Identify trends and patterns");
    println!("   6.  Synthesize comprehensive overview");
    println!("   7.  Provide detailed citations and references");

    // Demonstrate search refinement
    let refined_search =
        search_vector_store_with_limit("research-db-123", "algorithmic bias mitigation", 20)
            .filter("category", "methodology")
            .filter("confidence", "high");

    println!("\n Search Refinement:");
    println!("   Query: algorithmic bias mitigation");
    println!("   Limit: {} results", refined_search.limit_ref().unwrap());
    println!("   Filters: category=methodology, confidence=high");

    println!("\n Expected Research Response:");
    println!("    Executive Summary:");
    println!("      • Overview of current bias mitigation approaches");
    println!("      • Key methodological categories identified");
    println!("      • Emerging trends and best practices");
    println!("   ");
    println!("    Detailed Analysis:");
    println!("      • Pre-processing techniques (data augmentation, sampling)");
    println!("      • In-processing methods (fairness constraints, adversarial training)");
    println!("      • Post-processing approaches (threshold optimization, calibration)");
    println!("   ");
    println!("    Comprehensive Citations:");
    println!("      • [Smith et al., 2023] - Fairness constraints in ML training");
    println!("      • [Johnson & Lee, 2024] - Bias detection in neural networks");
    println!("      • [Research Survey, 2024] - Comprehensive bias mitigation review");
    println!("   ");
    println!("    Future Research Directions:");
    println!("      • Intersectional bias analysis");
    println!("      • Real-time bias monitoring");
    println!("      • Domain-specific mitigation strategies");

    Ok(())
}

pub fn filter_ref(&self) -> &HashMap<String, String>

Get the search filters.

Examples found in repository

examples/vector_stores.rs (line 275)

fn run_semantic_search_example() -> Result<(), Error> {
    println!("\n Example 3: Semantic Search and Similarity Queries");
    println!("{}", "=".repeat(60));

    // Create a search-optimized vector store
    let search_store = simple_vector_store("Semantic Search Demo Store")
        .add_file("file-ml-concepts-001")
        .add_file("file-nlp-techniques-002")
        .add_file("file-deep-learning-003")
        .add_file("file-computer-vision-004")
        .add_file("file-ai-ethics-005")
        .metadata("domain", "machine_learning")
        .metadata("search_optimized", "true");

    println!(" Created search-optimized vector store:");
    println!("   Name: {}", search_store.name_ref().unwrap());
    println!("   Domain: Machine Learning");
    println!("   Documents: {} files", search_store.file_count());

    // Demonstrate various search patterns
    println!("\n Search Query Examples:");

    // Basic semantic search
    let basic_search = search_vector_store("search-store-123", "neural network architectures");
    println!("   1. Basic Search:");
    println!("      Query: '{}'", basic_search.query());
    println!("      Store: {}", basic_search.vector_store_id());

    // Limited result search
    let limited_search = search_vector_store_with_limit(
        "search-store-123",
        "natural language processing techniques",
        5,
    );
    println!("   2. Limited Results:");
    println!("      Query: '{}'", limited_search.query());
    println!(
        "      Limit: {} results",
        limited_search.limit_ref().unwrap()
    );

    // Advanced filtered search
    let filtered_search =
        search_vector_store_with_limit("search-store-123", "computer vision applications", 10)
            .filter("category", "practical_applications")
            .filter("difficulty", "intermediate");

    println!("   3. Filtered Search:");
    println!("      Query: '{}'", filtered_search.query());
    println!(
        "      Filters: {} applied",
        filtered_search.filter_ref().len()
    );
    for (key, value) in filtered_search.filter_ref() {
        println!("         {}={}", key, value);
    }

    // Demonstrate search result processing
    println!("\n Search Result Processing:");
    println!("    Semantic similarity ranking");
    println!("    Document excerpt extraction");
    println!("    Relevance score calculation");
    println!("    Source location identification");
    println!("    Related content suggestions");

    // Show different query types
    println!("\n Query Type Examples:");
    let query_examples = vec![
        (
            "Conceptual",
            "What is machine learning?",
            "Broad conceptual understanding",
        ),
        (
            "Technical",
            "How to implement backpropagation?",
            "Specific technical implementation",
        ),
        (
            "Comparative",
            "LSTM vs Transformer architectures",
            "Comparative analysis",
        ),
        (
            "Problem-solving",
            "Overfitting in neural networks",
            "Problem identification and solutions",
        ),
        (
            "Application",
            "Computer vision in healthcare",
            "Domain-specific applications",
        ),
    ];

    for (query_type, query, description) in query_examples {
        println!("    {}: '{}'", query_type, query);
        println!("      Purpose: {}", description);
    }

    Ok(())
}

Trait Implementations

impl Clone for VectorStoreSearchBuilder

fn clone(&self) -> VectorStoreSearchBuilder

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for VectorStoreSearchBuilder

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.