BotMan AI maps legacy code with knowledge graphs

BotMan AI's pitch rests on the idea that modern AI models are capable enough, but they lack the business context needed to rebuild complex enterprise systems. Founder and CTO Alejandro D’Andrea describes a pipeline that ingests an entire legacy application at once, maps its dependencies, business rules and data models into a knowledge graph, and then rebuilds the application's architecture rather than merely translating old code line by line.

In his telling, the same graph can serve other tasks beyond migration. When a document arrives, relationships to projects, losses or lawsuits can be recorded once, and any agent later assigned a task can use that graph to find relevant information automatically. Linking unstructured information through entities — a person, project or product — is another way to recover context that would otherwise stay buried in separate messages and documents.

The interview was part of theCUBE + NYSE Wired: AI Luminaries, a series sponsored by Neo4j, with theCUBE acting as a paid media partner. That sponsorship is disclosed, and theCUBE says sponsors have no editorial control over its content.