Graph Query Startup PuppyGraph Raises $10M Funding to Let AI Agents Query 3,000 Tables Without a Graph Database

PuppyGraph, the Santa Clara-based graph query engine for enterprise AI agents, has raised $10 million in Series A funding led by Particle Future and defy.vc, bringing total funding to $15 million following a $5 million seed round led by defy.vc in November 2024 . The company did not disclose its valuation.

The funding comes after the general availability of PuppyGraph 1.0, which lets AI agents query enterprise data as a single unified graph—across up to 3,000 tables—without loading or copying data into a separate graph database. The company reports 7x annual recurring revenue growth over the past year, with customers including AMD, Palo Alto Networks, Datadog, Coinbase, and Trend Micro.

PuppyGraph Funding Overview By CREDX Media

Startup Name	PuppyGraph
Founders	Weimo Liu, Danfeng Xu, Lei Huang
Founded	2023
HQ	Santa Clara, California, USA
Round	Series A Funding
Amount Raised	$10M
Valuation	Not disclosed
Category	Graph Query Engine, Enterprise AI, Data Infrastructure
Lead Investors	Particle Future, defy.vc
Total Funding Raised	$15M (including $5M seed)
ARR Growth	7x year-over-year
Key Customers	AMD, Palo Alto Networks, Datadog, Coinbase, Trend Micro
Key Performance	10-hop query across billions of edges in 2.26 seconds

💡 Use of Funds

• Engineering Expansion: Deepening the distributed graph query engine and AI agent integration capabilities .
• Enterprise Go-to-Market: Scaling sales and marketing to reach more large enterprises building AI agent applications .

Startup Overview: PuppyGraph

What PuppyGraph Is Building in Simple Words

PuppyGraph has built a graph query engine that lets AI agents query existing enterprise data as if it were a graph—without moving that data into a dedicated graph database. The problem it solves: a proof-of-concept AI agent might work across 30 to 50 tables, but in production, agents need to navigate relationships across thousands of tables to answer complex questions like “If this account is compromised, which critical systems could an attacker reach?” .

The platform connects directly to existing data sources—Apache Iceberg, Databricks, Snowflake, ClickHouse, DuckDB, and PostgreSQL—and treats them as a unified graph. Agents can query relationships in plain English through PuppyGraph’s built-in AI agent or compatible tools via Model Context Protocol (MCP) .

• Zero-ETL Graph Queries: No data loading, no pipelines, no second copy of data .
• AI-Assisted Schema Builder: Automatically generates a graph schema from existing tables for teams to review and approve, reducing manual modeling .
• Ontology Validation: Rejects undefined relationship references and returns feedback agents can use to correct their queries—preventing agents from inventing connections that don’t exist .
• Enterprise Security: Row-level permissions enforced throughout graph traversals, single sign-on, role-based controls, and credential vending for short-lived identity-aware credentials .

Why PuppyGraph Matters

• Agents hit a wall at production scale: Schema descriptions consume context windows, join paths multiply, and deep relationship queries time out. PuppyGraph’s distributed engine completed a 10-hop query across billions of edges in 2.26 seconds on a four-machine cluster .
• The graph database migration problem is real: Moving data into Neo4j or TigerGraph introduces loading delays, pipeline maintenance, and scalability limits. PuppyGraph lets enterprises keep their existing data platform and query it as a graph in place .
• Customers are already running agents at scale: Sola Security runs AI agents across thousands of tables and billions of edges daily. Coinbase connects directly to Snowflake and Databricks tables—”standing up a new graph is a schema definition, not a data migration” .
• The graph market is growing: Fortune Business Insights estimates the graph database market will grow from $2.85 billion in 2025 to $20.29 billion by 2034 .

Founder Background: The Google and LinkedIn Infrastructure Veterans

Weimo Liu (CEO) worked on Google’s F1 team—the unified SQL engine serving billions of queries per day—and at TigerGraph, where he built query languages for distributed graph databases. He holds a PhD in Computer Science from George Washington University .

Danfeng Xu spent nine years on LinkedIn’s infrastructure team .

Lei Huang is a three-time Google Code Jam world finalist .

Liu’s framing of the opportunity: “The insights of enterprise data will soon be generated by agents. Users will expect to ask questions that cross departments and systems, beyond the workflows someone built into a dashboard. That requires agents to navigate thousands of tables, follow the right relationships, and retrieve evidence in subseconds” .

Founder Intelligence: What Founders Can Learn from PuppyGraph

• Solve the migration problem, not the database problem: Neo4j and TigerGraph require moving data. PuppyGraph’s insight was that enterprises don’t want another database—they want graph queries on the data they already have.
• Let agents correct themselves: PuppyGraph’s ontology validation rejects undefined relationships and returns structured feedback. Instead of hallucinating connections, agents can inspect the model and fix their queries.
• Performance is a product feature: In agent workflows, every hypothesis requires queries. A query that takes minutes interrupts the entire investigation. PuppyGraph optimized for subsecond traversals because that’s what interactive agents need.
• Build on open standards: MCP integration lets enterprises use the AI tools they already trust—Claude, ChatGPT—rather than forcing them into PuppyGraph’s agent.

Investor Intelligence: Why This Round Gets Premium Capital

Particle Future and defy.vc co-led the round. defy.vc also led PuppyGraph’s seed round, signaling continued conviction.

Amy Yin, Partner at defy.vc, framed the thesis: “Big data is still an unsolved problem in the world of AI. You can ask Claude to write a query, but executing it across billions of relationships takes serious engineering” .

The investor composition signals conviction: defy.vc‘s doubled-down bet after seed, and Particle Future’s participation, validate PuppyGraph’s approach to the enterprise AI data layer.

The Road Ahead for PuppyGraph

Expansion Targets

With $10 million in capital, PuppyGraph will:

• Expand engineering to deepen the graph query engine and AI agent capabilities .
• Scale enterprise go-to-market to reach more companies building production AI agents .

The Competitive Landscape

PuppyGraph competes with established graph database vendors:

• Neo4j was valued at over $2 billion after a $325 million Series F in 2021.
• TigerGraph, where Liu once worked, is built around storing data in its own graph database .

PuppyGraph’s differentiation is that it doesn’t require storing data in a new system. For enterprises already running Snowflake, Databricks, or Iceberg, that removes a major adoption barrier.

Conclusion

PuppyGraph’s $10 million Series A funding validates the thesis that AI agents need a graph query layer—but not another graph database. The company’s zero-ETL approach, 7x ARR growth, and customers including AMD and Coinbase demonstrate that enterprises want graph intelligence without data migration.

Market tailwinds favor PuppyGraph: the explosion of AI agent deployments, the complexity of enterprise data relationships, and the growing recognition that agents need structured business context to avoid hallucinating connections. The company’s Google and LinkedIn pedigree, proven performance, and customer traction position it to capture the data layer for enterprise AI agents.

For deep analysis of AI infrastructure funding, graph technology innovation, and companies reshaping how agents query enterprise data, explore CREDX Media for comprehensive coverage of startup funding trends, valuation intelligence, and founder playbooks.

Want the latest funding news and market insights delivered straight to your inbox? Subscribe to CredX Letters for verified deal terms, valuation benchmarks, and daily macro signals.

Frequently Asked Questions

1. What is PuppyGraph’s total funding and latest round size?
PuppyGraph has raised $15 million total. Its $10 million Series A was led by Particle Future and defy.vc in November 2026, following a $5 million seed round in November 2024 .

2. How much did PuppyGraph raise in its Series A round and who led the investment?
PuppyGraph raised $10 million co-led by Particle Future and defy.vc .

3. What is the valuation of PuppyGraph?
PuppyGraph did not disclose its post-money valuation.

4. Who are the major investors in PuppyGraph?
Major investors include Particle Future and defy.vc (co-leads). defy.vc also led the seed round .

5. Who are PuppyGraph’s founders?
PuppyGraph was founded in 2023 by Weimo Liu (CEO), Danfeng Xu, and Lei Huang. Liu worked on Google’s F1 team and at TigerGraph; Xu spent nine years at LinkedIn .

6. What is PuppyGraph’s business model?
PuppyGraph operates a B2B SaaS model, selling its graph query engine to enterprises building AI agent applications. A free developer version is also available via Docker .

7. How does PuppyGraph’s technology work?
PuppyGraph connects directly to existing data platforms and queries them as a unified graph without moving data. An AI-assisted schema builder generates the graph model, and agents can query it via MCP or PuppyGraph’s built-in agent .

8. What traction justifies PuppyGraph’s Series A funding?
PuppyGraph’s ARR grew 7x over the past year. Customers include AMD, Palo Alto Networks, Datadog, Coinbase, and Trend Micro .

9. What are PuppyGraph’s plans for the $10 million Series A capital?
PuppyGraph will expand engineering and scale enterprise go-to-market .

10. Who are PuppyGraph’s main competitors?
Competitors include Neo4j (valued at $2B+) and TigerGraph, both of which require storing data in their own graph databases .