AWS First Steps
LanguageEnglish
Database

Building a Graph Database with Amazon Neptune - Knowledge Graphs and Social Network Analysis

Manipulate graph data with two query languages, Gremlin and SPARQL, and achieve query scaling with up to 15 read replicas. Learn how to use Neptune Analytics for graph algorithm execution and vector search.

About 5 min read

Overview of Neptune and Graph Database Benefits

Amazon Neptune is a fully managed graph database service. It supports two graph models: property graphs (Gremlin/openCypher) and RDF graphs (SPARQL), suited for graph-based use cases such as social networks, recommendations, knowledge graphs, and fraud detection. Queries that require multi-level JOINs in relational databases (such as finding friends of friends of friends) can be naturally expressed as graph traversals. It provides up to 15 read replicas for read scaling, multi-AZ deployment for high availability, and point-in-time recovery for data protection as standard. Storage automatically scales up to 128 TiB, with a high-availability architecture that maintains 6 data copies across 3 AZs.

Query Languages and Data Models

Gremlin is a traversal language for the property graph model, where vertices (nodes) and edges have properties (key-value pairs). A traversal like g.V().has("person","name","Alice").out("knows").values("name") retrieves the names of Alice"s friends. openCypher is a declarative query language originating from Neo4j, allowing pattern matching queries like MATCH (p:Person {name:"Alice"})-[:KNOWS]->(f) RETURN f.name. SPARQL is a W3C standard query language for RDF (Resource Description Framework) graphs, suited for building knowledge graphs and ontologies. You can use Gremlin/openCypher and SPARQL simultaneously within the same cluster, but the data models are managed separately. Neptune ML applies machine learning to graph data, executing tasks such as node classification and link prediction with GNN (Graph Neural Networks).

Practical Use Cases

Graph databases have a wide variety of use cases. In fraud detection, graph algorithms detect anomalous patterns within transaction networks (circular transactions, collusion networks). Recommendation engines integrate user purchase history, browsing history, and social graphs for recommendations. In IT infrastructure dependency management, the relationships between servers, applications, and network devices are visualized as graphs, making it easier to identify the failure blast radius. In life sciences, protein interaction networks and drug side-effect relationships are analyzed via graphs for drug discovery research.

Neptune Analytics is a serverless analytics engine that executes graph algorithms (PageRank, shortest path, community detection, centrality analysis) on graph data. It ingests data from Neptune Database snapshots or S3, enabling interactive algorithm execution on graphs. Vector search is also integrated, allowing you to store vector embeddings on graph nodes and combine similar node searches with graph traversals. For example, you can attach text embeddings to knowledge graph entities and build a RAG pipeline that integrates semantic search with graph exploration.

Hybrid Architecture with DynamoDB and Data Pipelines

Combining Neptune with DynamoDB enables an architecture that balances graph query flexibility with DynamoDB"s fast reads and writes. You can build a hybrid configuration storing entity attribute data in DynamoDB and inter-entity relationships in Neptune. Lambda functions integrate both services, providing a unified API to clients via API Gateway. Neptune Streams detects graph data changes in real time, enabling event-driven pipelines that automatically update DynamoDB caches. Neptune"s bulk loader imports data such as CSV from S3. Step Functions can build graph data ETL pipelines to automate periodic data updates. Neptune Serverless automatically scales with workload, keeping costs down during idle periods.

Neptune Pricing

Neptune Database pricing consists of instances, storage, and I/O. Instance pricing varies by instance class and Region. Neptune Serverless uses capacity unit (NCU)-based billing, auto-scaling from a minimum of 1 NCU to a maximum of 128 NCU. Storage and I/O are billed based on usage. For intermittent workloads, Serverless is suitable.

Summary

Amazon Neptune is a fully managed graph database supporting three query languages: Gremlin, openCypher, and SPARQL. It can be used for relationship analysis, including fraud detection, recommendations, knowledge graphs, and dependency management. With Neptune Analytics for graph algorithm execution and integrated vector search, plus hybrid architectures with DynamoDB, you can build advanced graph-based systems.

References (Official AWS Resources)

The primary sources for this page are the official AWS website and documentation. Check the official pages below for the latest specifications and pricing.

If this page and the official documentation disagree, treat the official documentation as authoritative.

Share

Related Services

Amazon Neptune SpecializedA managed graph database service for processing highly connected datasetsAmazon S3 EssentialA scalable object storage serviceAWS Lambda EssentialA serverless compute service that runs code without server management

Related Articles

Amazon DynamoDB Table Design Patterns - Single-Table Design and GSI StrategiesLearn about DynamoDB partition key design, single-table design, and how to implement diverse access patterns using GSIs.Running MongoDB Workloads as a Managed Service with Amazon DocumentDB - Document Model and Query DesignDocumentDB manages document data with a MongoDB-compatible API and achieves read scaling and DR with up to 15 read replicas and global clusters. This article explains how to use sharding with Elastic Clusters.Practicing Infrastructure as Code with AWS CloudFormation - Template Design and Stack ManagementDefine infrastructure with templates and preview impact with change sets before deployment. Detect configuration drift with drift detection and deploy across your entire organization with StackSets.

More on This Topic

Multi-Region Architectures with Amazon Aurora Global Database - Designing for DR and Global ReadsAchieves an RPO of less than 1 second through physical replication at the storage layer. Introduces the procedures for planned and unplanned failover and how to use Write Forwarding for global reads.Cassandra-Compatible Database - Serverless Distributed Database with Amazon KeyspacesLearn how to design and operate distributed databases using Amazon Keyspaces (for Apache Cassandra) and DynamoDB.Achieving Microsecond Latency with Amazon DynamoDB Accelerator (DAX) - In-Memory Cache DesignLearn about read acceleration for DynamoDB with DAX, cache strategies, and cluster design.Document Database in Practice - Flexible Data Modeling with Amazon DocumentDB and DynamoDBLearn how to design and operate document databases using Amazon DocumentDB and DynamoDB.Building Event-Driven Architectures with Amazon DocumentDB Change StreamsThis article introduces how to build event-driven architectures for real-time data synchronization by integrating DocumentDB's change data capture with Lambda triggers.Running MongoDB Workloads as a Managed Service with Amazon DocumentDB - Document Model and Query DesignDocumentDB manages document data with a MongoDB-compatible API and achieves read scaling and DR with up to 15 read replicas and global clusters. This article explains how to use sharding with Elastic Clusters.Running a MongoDB-Compatible Database with Amazon DocumentDB - Design Patterns and ScalingYou select an instance class for MongoDB-compatible DocumentDB and secure scalability with sharding through Elastic Clusters. This article also introduces backup strategies.Amazon DynamoDB Table Design Patterns - Single-Table Design and GSI StrategiesLearn about DynamoDB partition key design, single-table design, and how to implement diverse access patterns using GSIs.

Similar Articles and Services

Amazon NeptuneA fully managed graph database service supporting Gremlin, openCypher, and SPARQL that can traverse billions of relationships with low latencyAWS Purpose-Built Database Strategy - The Workload Optimization Philosophy Behind 15+ Specialized DatabasesWe explain the design philosophy behind AWS's 15+ purpose-built database services, comparing them with Azure Cosmos DB's unified approach and GCP's Cloud Spanner/Bigtable, and discuss the benefits of choosing the optimal database for each workload.Security Investigation and Threat Analysis - Streamlining Incident Response with Amazon DetectiveLearn how to use Amazon Detective for security incident investigation and threat analysis. This guide covers the detection-to-investigation workflow with GuardDuty integration and root cause identification through graph-based analysis.Investigating Security Incidents with Amazon Detective - Root Cause Identification Through Graph AnalysisLearn about investigating GuardDuty findings with Detective, entity profiles, and leveraging behavior graphs.