Zep

Dynamic Graph RAG

Fast graph retrieval and real-time updates for changing business data

Get Started FreeView Documentation

Why Zep for Graph RAG

Fast Retrieval and Quick Graph Updates

Traditional RAG works for static documents. Zep's Graph RAG handles dynamic business data that changes constantly: customer records, support cases, user preferences, and real-time interactions. Your agents access current business context without expensive batch recomputation.

Stylized graph visualization showing connected business knowledge nodes

How Zep's Graph RAG Compares

Semantic RAGMicrosoft GraphRAGZep
Use CaseStatic DocumentsStatic Corpus SummarizationReal-time, Dynamic Data
UpdatesBatch recomputationBatch recomputationReal-time incremental
Temporal HandlingFact conflicts are unresolvedLimited to LLM judgement during summarizationTracks fact validity over time
Query SpeedFast (ms)Slow (seconds)Fast (ms)

Real-Time Business Intelligence

Dynamic Data Integration

Add business data as JSON, text, or structured messages. Customer records, support tickets, transaction data, and user interactions automatically become part of your knowledge graph without rebuilding anything.

Incremental Updates

When business data changes (customer status updates, new support cases, preference changes), Zep updates the graph in real-time. No waiting for batch processes or expensive recomputation cycles.

Temporal Fact Management

Track how business information evolves over time. When a customer's status changes or a support case closes, Zep maintains historical context while surfacing current data. Your agents understand both what happened and when.

Relationship-Aware Retrieval

Find connected business information through entity relationships. When querying about a customer, automatically surface related support cases, purchase history, and team interactions without manual linking.

Graphs for Shared Knowledge

Cross-User Business Context

Create graphs for shared business knowledge like product information, company policies, and team resources that multiple users and agents can access. Avoid duplicating the same business data across individual user graphs.

Custom Entity Types

Define business-specific entities like customers, products, cases, or projects. Model relationships that matter to your domain: purchases, support interactions, team assignments, or project dependencies.

Flexible Data Sources

Ingest from CRM systems, support platforms, databases, or APIs. Zep handles JSON documents, unstructured text, and structured business records in the same graph.

Developer Experience

Simple Integration

Add business data with a single API call. Search across user and graphs with the same interface. Works with your existing LLM frameworks and applications.

Performance at Scale

Millisecond query responses even with large business datasets. Hybrid search combines semantic similarity, keyword matching, and graph traversal for comprehensive results.

Custom Search Filters

Filter by entity types, date ranges, or custom attributes. Find specific business relationships or track changes over time periods that matter to your use case.

graph_rag_example.py
# Add business data to graph
client.graph.add(
    group_id="product_team",
    type="json", 
    data=json.dumps(customer_data)
)

# Search across business context
results = client.graph.search(
    group_id="product_team",
    query="customer support issues",
    scope="edges"
)

Customized for Your Domain

Zep adapts to your business through custom entity types and relationship models. These models enable precision recall of exactly the required context, so your agents understand your business domain, not just generic conversations.

Sales and Marketing

Store lead preferences, product interests, campaign interactions, and buying signals.

Your sales agents understand prospect history, pricing discussions, and engagement patterns to personalize outreach and close deals faster.

sales_and_marketing_entities.py
class Lead(EntityModel):
    """Represents a sales lead or prospect."""
    company_size = Field(
        description="startup, SMB, mid-market, enterprise"
    )
    budget_range = Field(
        description="Budget discussed or indicated"
    )
    decision_timeline = Field(
        description="Expected decision timeframe"
    )

Ready to Connect Your Business Data?

Deploy Graph RAG that understands your dynamic business context, not just static documents.

Start Building
Enterprise-ready
SOC2 Type 2 certified • HIPAA compliant • Flexible deployment options