TigerGraph Savanna

TigerGraph lists offices in Milpitas, California and in the United Kingdom, and builds a native, massively parallel graph database for enterprises that need to reason over billions of relationships in real time, for example in fraud detection, customer 360, supply chain and network analysis.

The platform stores data as a graph and runs multi-hop queries across petabyte-scale datasets in milliseconds, with GSQL, openCypher and GQL query languages, hybrid graph and vector search, GraphRAG for agentic AI and a graph data science library. It is offered as TigerGraph Savanna, a fully managed cloud service with bring-your-own-cloud and key options, as self-managed TigerGraph Enterprise on Kubernetes, AWS, Azure, GCP or on-premises, and as a free single-server Community Edition. Customers named on the site include JP Morgan Chase, Mastercard, American Express, Intuit, AT&T, Ford, Xbox and Jaguar Land Rover.

• Massively parallel graph engine: Native graph storage for deep multi-hop queries on billions of relationships.
• Hybrid graph and vector search: Combines relationship traversal with vector similarity for GraphRAG.
• Query languages: GSQL, openCypher and GQL with GraphStudio and Insights tools.
• Enterprise security: Mutual TLS, Kerberos, SAML, OIDC and LDAP single sign-on and role-based controls.
• Lakehouse integration: Apache Iceberg connector, Kafka, Spark and export to S3-compatible storage.
• Savanna cloud: Fully managed workspaces with automatic backups, rolling upgrades and BYOC options.

A Vietnamese bank or e-wallet can score transfers in real time by checking links to known mule accounts and shared devices. A telecom operator can detect SIM fraud rings and analyze network dependencies. A retailer can build customer 360 profiles by resolving identities across loyalty, online and store data.

Investigators and data teams can follow relationships across 10 or more hops in milliseconds, which the site links to 300 percent faster fraud investigations and a 3x increase in identity resolution rates. Intuit reports detecting 50 percent more risk events with 50 percent better model precision. Usage-based Savanna pricing and a free Community Edition supporting up to 300 GB let teams start small before scaling past 100 TB on Enterprise.

Fraud, anti-money laundering and AI agents increasingly depend on understanding relationships between accounts, devices, people and transactions rather than isolated records. Southeast Asian banks, e-wallets and telecom operators face rising scam and mule-account activity, creating demand for graph analytics that can run at national transaction volumes.

TigerGraph lists its first office at 680 N McCarthy Blvd, Milpitas, California, plus a UK office. Savanna runs in US, European and Asia-Pacific regions including Singapore, Sydney and Tokyo. TigerGraph DB 4.3, released in July 2026, added mutual TLS, Kerberos SSO and an Apache Iceberg connector.