PuppyGraph Secures $10M Series A After Sevenfold ARR Growth

PuppyGraph announced on November 10, 2026 that it raised a $10 million Series A led by Particle Future, with participation from defy.vc, to expand engineering and enterprise go-to-market work. The financing brings the graph query engine developer’s total funding to $15 million.

The round follows the general availability of PuppyGraph 1.0. PuppyGraph reported that annual recurring revenue grew sevenfold over the past year, with customers including AMD, Palo Alto Networks, Coinbase, Trend Micro, Datadog and Sola Security.

Querying Enterprise Data in Place

PuppyGraph is aimed at AI agents that investigate business questions across departments rather than simply generate SQL for a known dataset. An agent may need to follow relationships among customers, products, components and suppliers; as tables multiply, schema descriptions consume more of its context window and possible join paths increase.

The engine applies graph computation directly to supported databases, warehouses and data lakes. Records stay in systems such as Apache Iceberg, Databricks, Snowflake, PostgreSQL and Microsoft Fabric, avoiding a dedicated graph database, a separate graph copy and the associated loading pipelines. Applications can still use SQL for filters and aggregations while choosing graph queries for relationship-heavy questions.

An ontology supplies approved business entities and relationships, giving agents valid investigation paths. PuppyGraph’s AI-assisted schema builder proposes that model from existing tables, but data teams review and approve its business meaning. Compatible AI tools connect through MCP, and users can also work through PuppyGraph’s built-in agent.

The engine rejects references to entities or relationships absent from the ontology and returns structured feedback for query correction. That addresses silent errors caused by an agent inventing a connection and treating the result as evidence. PuppyGraph cautions that validation does not guarantee a correct conclusion because an agent can still pursue a poor hypothesis or conduct an incomplete analysis.

Performance and Governance Controls

PuppyGraph reported that its engine supports subsecond graph queries for supported workloads and enables agents to query across 3,000 enterprise tables as one graph. In a separate company benchmark, a four-machine cluster completed a 10-hop query across billions of edges in 2.26 seconds, with performance dependent on workload.

Speed affects how investigations unfold: agents test multiple hypotheses, and a slow query can stall the next analytical step when it depends on the previous result. The benchmark therefore bears on whether deep relationship searches can remain interactive, rather than serving only as a measure of raw query throughput.

PuppyGraph 1.0 also enforces configured row-level permissions throughout graph traversals, including intermediate steps. Role-based controls and user impersonation tie access to user identity. Single sign-on is supported, while administrative audit logs record management activity. Credential vending obtains short-lived, identity-aware credentials for platforms such as Snowflake and Databricks without requiring users to expose or manage long-lived credentials.

Deployments and Expansion Priorities

Sola Security CTO Leon Goldberg said: “We run AI agents across thousands of tables and billions of edges a day.” Palo Alto Networks engineering director Ran Xia said the company runs multi-hop traversals on its existing data platform without moving or duplicating rows, eliminating the associated ETL pipeline maintenance. Coinbase senior staff software engineer Xinyu Liu said Coinbase connects directly to tables maintained in Snowflake and Databricks and uses MCP for natural-language relationship queries.

AMD began with IT workloads and is expanding toward information-security and digital-marketing data, PuppyGraph reported. That expansion illustrates the company’s broader model: as a relationship map connects additional departments, agents can investigate questions that previously stopped at organizational boundaries.

The company plans to use the new capital to reduce graph-model setup work, make relationship queries practical at enterprise scale and help customers move agents into production. The central challenge is to broaden those investigations without weakening the human-approved ontology and permission controls that determine which relationships are valid and which data each user can see.