Blitzy Bets on Neo4j Knowledge Graphs to Ground Autonomous Coding Agents

In an interview with theCUBE Research’s John Furrier during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio, Deshmukh described the central obstacle for autonomous software development. “I think when people talk about changing software autonomously, the biggest problem is not whether AI can write code. It’s about does AI know or understand the system that it’s writing the code for?” he said. “The reason that understanding is crucial is because I may be changing a line of code here, but it may have downstream impact somewhere far away.”

Blitzy raised $200 million at a $1.4 billion valuation in May to run thousands of coding agents in parallel, according to SiliconANGLE. Its platform starts by reverse-engineering a customer’s existing environment and building a dynamic graph of its codebase. That graph plugs into source control systems such as GitHub and GitLab, so it updates whenever developers or agents change the code. Deshmukh said using knowledge graphs reflects a basic fact about software. “Fundamentally, code is a graph because you think about, ‘Oh, we have modules, we have files, we have functions, we have objects, we have classes, variables,’” he said. “Each of these are entities that are related to each other. So even before you bring AI into the picture, a codebase is a graph intrinsically.”

Without that structure, agents fall back on vector searches or grep commands to trace dependencies, and those searches burn through working memory quickly. Deshmukh noted that an agent’s effective context tops out at about 200,000 to 300,000 tokens, or about 20,000 to 30,000 lines of code. On a 100-million-line codebase, agents start compacting results and losing information. “With a graph, you know exactly what you’re going to reach because you basically get to pick the point where you want to start. And you know exactly what is accessible. And so you have that effective context. Every agent knows the context that it needs to and nothing else,” he said.

That efficiency changes how projects get scoped. With less context lost to broad searches, Blitzy can tackle whole projects at once rather than splitting them into the epics, user stories and tasks of the traditional sprint model. Quality control is built into the process: humans approve an Agent Action Plan before coding begins, and every line of generated code is tested immediately. Blitzy cited an 84.95% score on SWE-Bench Pro in June. “On top of that, we have agents who are watching other agents to make sure that they abide by the spec, the plan that we had created and approved by the human in the loop,” Deshmukh said. “That ensures that there is no drift or hallucination.”

Query language also matters. Neo4j used GraphSummit to pitch knowledge graphs as shared context for AI agents, and Blitzy’s agents run queries in Neo4j’s Cypher language constantly. The strictness of Cypher doubles as a safeguard because a malformed query from a hallucinating agent simply returns nothing, Deshmukh said. “You only get a result for a correct query,” he said. “So there is no question of the agents working off of made-up information or something false. You’re always grounded in truth of what’s in the knowledge graph.”

The video interview was part of SiliconANGLE’s and theCUBE’s coverage of GraphSummit. TheCUBE is a paid media partner for GraphSummit. Neither Neo4j, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.