PuppyGraph Raises $10M Series A as Its 1.0 Graph Query Engine Adds an AI-Assisted Schema Builder and MCP Integrations
PuppyGraph, which queries warehouse and lakehouse data as a graph, said it raised $10 million and added an AI-assisted schema builder and MCP integrations; an analyst says the graph model still needs upkeep.
Overview
PuppyGraph, a startup whose query engine lets companies treat data already sitting in warehouses and lakehouses as a single graph, said on November 10, 2026 that it has raised $10 million in new funding, according to SiliconANGLE. On the same day, InfoWorld reported that the company said it was adding an AI-assisted graph schema builder and MCP integrations to PuppyGraph 1.0, the new version of its distributed graph query engine. SiliconANGLE reports that the round follows the general release of version 1.0; neither article gives a date for that release.
The funding
SiliconANGLE says the money is meant to build out PuppyGraph’s enterprise go-to-market operation, with some of it also going into engineering. Particle Future and defy.vc led the Series A round. It is the second time defy.vc has led a round for the company, after a $5 million seed round in November 2024, and investors have now put $15 million into the startup, the outlet reports.
The outlet also reports that PuppyGraph’s annual recurring revenue grew sevenfold over the past year “by the startup’s count”, and notes that no dollar figure was disclosed. Weimo Liu, co-founder and chief executive, is quoted as saying: “The insights of enterprise data will soon be generated by agents”.
What the release changes
According to InfoWorld, the engine maps existing tables into a Labeled Property Graph model, with data entities as nodes, relationships as edges and attributes as properties, and also supports conventional SQL queries. Among the changes InfoWorld and SiliconANGLE describe, as stated by the company:
- Schema builder. The AI-assisted builder automatically maps existing tables into graph entities, relationships and properties, and developers review and approve the result instead of defining it by hand, InfoWorld reports.
- Query validation. The engine validates agent-generated queries against the graph schema, rejecting undefined relationships and returning feedback that agents can use to correct errors, per InfoWorld. SiliconANGLE describes the same behaviour and says the error is written for the agent to act on.
- Distributed execution. Complex graph queries are split into smaller operations that run simultaneously across multiple compute nodes, according to InfoWorld, which attributes this to the company.
- MCP integrations. These let teams connect PuppyGraph to AI agents and tools that support the Model Context Protocol, InfoWorld says. SiliconANGLE adds that PuppyGraph also has its own agent for asking questions in plain English.
- Credentials and permissions. When PuppyGraph connects to Snowflake or Databricks it can obtain short-lived credentials tied to the user’s identity, and the company says it checks row-level permissions at each hop of a traversal, according to SiliconANGLE.
SiliconANGLE adds that developers can download a free version with a single Docker command and that the company says it takes about 10 minutes to go from deployment to a first query.
Figures that come from the company
SiliconANGLE reports that the company says agents running on PuppyGraph can work across 3,000 enterprise tables and get graph query results back in under a second. The outlet also relays the company’s argument that a proof of concept might cover 30 to 50 tables, and that teams scaling past that find schema descriptions alone eating into an agent’s context window. The SiliconANGLE article does not describe the test conditions behind the 3,000-table and sub-second figures, the InfoWorld article does not carry them, and neither article reports an independent benchmark.
SiliconANGLE says PuppyGraph’s customer list includes AMD and Palo Alto Networks and that the announcement also named Coinbase, Datadog and Trend Micro as customers. InfoWorld says PuppyGraph stated that its technology is already used by enterprises including AMD, Palo Alto Networks, Datadog, Netskope, Coinbase, Trend Micro, Baker Hughes, Sola Security, HighStreet Insurance Partners and Prevalent AI.
Analyst views reported by InfoWorld
InfoWorld names three analysts. Devin Pratt, research director at IDC, said the features can reduce the engineering effort involved in connecting agents to enterprise data, and is quoted as saying: “Developers model entities and relationships once, then query connected data with graph languages instead of writing and maintaining long chains of joins”. Ashish Chaturvedi, executive research leader at HFS Research, pointed to what he terms PuppyGraph’s enforced ontology as the bigger gain for agentic workflows.
The same article sets out trade-offs. Michael Leone, principal analyst at Moor Strategy and Insights, said developers will still need to define and maintain the graph model, adding: “In other words, the complexity does not disappear; it shifts from writing and maintaining long SQL joins to designing and governing the graph model”. Pratt echoed that query quality will depend on the consistency of the underlying schemas and the quality of the source data. Chaturvedi cautioned: “Because PuppyGraph queries data in warehouses and lakehouses, running graph-intensive workloads can increase compute consumption on those systems, requiring CIOs to model the costs before scaling deployments”.
An analyst quoted in the article also said the engine is primarily built for analytical queries rather than applications that require low-latency graph writes or high volumes of transactional updates, which may still be better suited to native graph databases, and that CIOs should consider the maturity and support capacity of a smaller vendor.
What the cited articles do not say
Neither article gives pricing beyond the free download, the date of the 1.0 general release, or which specific agents and tools the MCP integrations support.