Cogent Security, an applied AI company building autonomous agents that automate the detection and remediation of cybersecurity vulnerabilities, has closed a $42 million Series A round led by Bain Capital Ventures, with participation from existing investor Greylock Partners and Definition. The raise brings Cogent's total funding to $53 million, only six months after the company officially launched.
The company was built to solve a specific and growing problem: security teams can find vulnerabilities, but fixing them is slow, manual, and coordination-heavy. Cogent's agents handle that entire process autonomously, from identifying who owns an affected system to tracking the fix through to verified completion.
The numbers tell a stark story. In 2025, more than 48,000 new software vulnerabilities were published, 162 percent higher than five years ago. New vulnerabilities appear constantly as code changes and systems evolve, and the time it takes hackers to weaponize those flaws has collapsed to minutes.
Defenders, by contrast, still work in days and weeks. Security teams hunt down system owners through outdated directories and Slack threads. Developers receive tickets without knowing which of their systems are affected or why one issue matters more than the hundreds sitting in their backlog. Finding vulnerabilities is no longer the hard part. Modern scanning tools handle that. The real bottleneck is everything that follows: determining ownership, assessing actual risk, coordinating fixes, and proving the work got done.
Traditional security tools stop at detection. They surface a finding and hand the rest off to humans. Cogent takes a different approach. Its AI agents investigate vulnerabilities, identify who owns the affected systems, assess real-world risk based on the specific environment, and produce remediation steps ready for engineers to act on. The agents then track each fix through to verified completion.
Customers using Cogent have reduced their exposure windows for critical risks by an average of 97 percent. Every agent action is traceable and reproducible. The platform integrates directly into existing security and engineering workflows and operates under configurable approval gates and policy enforcement. Organizations can deploy autonomous agents without giving up control or auditability.
"Security teams are drowning in coordination work: chasing down system owners, writing tickets, proving fixes happened. We built AI agents that handle that work end-to-end, so security teams can finally keep pace with attackers," said Vineet Edupuganti, CEO and co-founder of Cogent.
With the new capital, Cogent plans to accelerate product development, expand enterprise deployments, and continue building agentic security infrastructure for large-scale, governed execution. Since launching in July 2025, the company is already working with dozens of Fortune 1000 companies. The goal is to move well beyond vulnerability detection and handle the full remediation lifecycle, from discovery to verified fix, without human bottlenecks at every step.
Cogent was founded by people who have spent their careers on both sides of the vulnerability problem. CEO and co-founder Vineet Edupuganti previously led product strategy at Abnormal Security. CTO and co-founder Geng Sng built the ML fraud detection system now protecting half of the Fortune 500, also at Abnormal Security. Co-founder Thanos Baskous led infrastructure at Coinbase, where he ran large-scale vulnerability remediation programs.
The team also includes AI researchers from Google DeepMind who contributed to the transformer architectures behind models like ChatGPT, Claude, and Gemini. That combination of security operations experience and foundational AI research is what sets the company apart technically.
Bain Capital Ventures led the round with a clear conviction that autonomous agents will define the next generation of enterprise security. The firm has a strong track record backing infrastructure and security companies, and sees Cogent as addressing a problem that has grown too large for traditional tooling. "The volume of vulnerabilities has outpaced what traditional, human-driven coordination can handle," said Enrique Salem, partner at Bain Capital Ventures. "Cogent's AI enables teams to accomplish five times more with the same resources."
Greylock Partners, which backed Cogent from the start, doubled down in this round. Its continued support signals confidence in both the team and the scale of the opportunity. "Cogent represents a new generation of security companies built natively for AI agents as first-class operators," said Saam Motamedi, partner at Greylock. "The technical depth of the team combined with that execution speed makes the customer signal impossible to ignore."
Definition also participated, alongside a notable group of individual investors. Founders and senior executives from OpenAI, Abnormal Security, and Datadog invested personally in the round. That kind of personal commitment from people who have built and scaled security and AI companies carries real weight.
