G5 Labs, an MIT spinout, just closed a $14 million seed round led by Pillar VC and Battery Ventures. The company's core idea is replacing traditional code with natural language organized as a formal graph—essentially making English the source code itself. Tim Kraska, the CEO and MIT professor, frames this as solving a real problem: teams are either struggling to adopt AI or drowning in thousands of lines of auto-generated code they can't review or maintain. He's positioning this as a fundamental shift in abstraction layers, similar to how high-level languages once replaced assembly code. The technology hinges on a bidirectional compiler that converts natural language to code and back based on what they call an ontology graph. The platform lets organizations manage intent across teams, resolve conflicts at the semantic level rather than line-by-line, trace code back to its original intent, and enforce policies like GDPR compliance automatically. In a real customer project at a financial services firm, the system caught structural problems during a legacy modernization that would've been invisible otherwise—and let business analysts drive design decisions directly instead of waiting for developers to translate requirements. The team includes veterans from AWS and Google plus MIT PhDs, with Kraska bringing twenty years of experience in ML and AI across academia and startups. Investors like Max Schireson (former MongoDB CEO) are betting that English will become the de facto language of software development, and G5 is building the engineering infrastructure to make that actually work at scale. Current customers are already using it to modernize complex systems in regulated industries without dragging forward decades of technical debt.