Defining the New Wave of Enterprise AI Agents
Enterprise AI agents are software systems that can plan, execute, and monitor complex tasks such as coding, database management, and infrastructure automation with limited human input, running continuously inside production environments and integrating tightly with existing tools, workflows, and data sources. The latest AI agent funding rounds for Cognition and Supabase show how quickly this model is moving from experiments to core business infrastructure. Cognition’s Devin, which positions itself as an autonomous AI software engineer, is now embedded in thousands of organizations’ development workflows. In parallel, Supabase is evolving into an agentic infrastructure backbone, where AI agents can provision, scale, and operate PostgreSQL databases. Together, these companies show that enterprise AI agents are no longer side projects or proofs-of-concept; they are beginning to reshape how teams build and run software at scale.
Cognition’s Billion-Dollar Bet on Devin as an AI Software Engineer
Cognition has raised more than USD 1 billion (approx. RM4.6 billion) at a USD 26 billion (approx. RM119.6 billion) valuation on the strength of Cognition Devin, its AI coding agent. Investors such as Lux Capital, General Catalyst, and 8VC are backing Devin’s rapid move into enterprise production environments. According to Cognition, enterprise usage of Devin has increased more than tenfold since the start of the year, pushing its annualized revenue run rate to USD 492 million (approx. RM2.3 billion). Customers include Citi, Mercedes-Benz, Goldman Sachs, Dell Technologies, Santander, the United States Army, and the United States Navy, along with fast-growing startups like Exa and Eight Sleep. These deployments are not cosmetic: Mercedes-Benz cut an eight-month modernization project down to eight days, and Itaú Unibanco now fixes 70% of security vulnerabilities automatically with Devin.
Supabase’s $500 Million Push into Agentic Infrastructure
Supabase has secured USD 500 million (approx. RM2.3 billion) in a Series F round at a USD 10.5 billion (approx. RM48.3 billion) valuation to build what it calls agentic infrastructure for AI-native applications. Led by GIC with participation from Accel, Y Combinator, Craft Ventures, and others, the round arrives only seven months after its previous financing. Supabase reports that its user base has more than doubled since the last round, while databases running on its platform have grown 600% year over year. The company now serves more than 250,000 customers and over 9 million developers, with AI agents increasingly responsible for provisioning and deploying databases. Its Supabase for Platforms product, which underpins many AI application builders, has seen 370% customer growth in six months. Supabase also introduced Multigres, an open source scaling layer that adds sharding, zero-downtime migrations, and high availability for PostgreSQL.
From Coding Desks to Datacenters: AI Agents Move into Production
The combined AI agent funding for Cognition and Supabase shows that investors now view enterprise AI agents as durable businesses rather than a passing model cycle. Cognition is moving DevOps and software engineering toward a “self-driving” pattern: engineers define problems, while Devin plans and executes the work. Internally, Cognition reports that 89% of its engineers’ committed code is now produced by Devin, a striking signal of how deeply AI agents can integrate into everyday workflows. Supabase, on the other hand, is turning its open source Postgres platform into an automation-friendly foundation where agents handle database provisioning, scaling, and failover. As AI agents span both application logic and infrastructure, enterprises gain end-to-end automation: coding, testing, deployment, and database management can all be orchestrated through agentic infrastructure rather than manual tickets and scripts.
