depthfirst, a company focused on applying AI to improve software security, has announced a major funding boost. They raised $80 million in a Series B round, with Meritech Capital leading the investment. Existing investors like Accel, Box Group, Liquid 2 Ventures, Alt Capital, and Mantis VC also participated, bringing the company’s total funding to $120 million. The timing of this raise is notable because it happened less than three months after their initial $40 million Series A funding, signaling strong investor confidence in both their product and the problem they are tackling.

Introducing depthfirst’s Proprietary AI Security Model

Along with the funding news, depthfirst introduced dfs-mini1, its first in-house AI security model. This new model is designed to protect cryptocurrency smart contracts initially. It was built using open-source tools and then further trained with reinforcement learning in security-specific environments. The company tested dfs-mini1 against OpenAI’s EVMBench, a standard benchmark for identifying vulnerabilities in smart contracts. Early results are promising, showing that dfs-mini1 outperforms other frontier models while operating at a fraction of the cost—10 to 30 times lower.

Internal tests also suggest that dfs-mini1 isn’t limited to smart contracts. Early signs indicate that the training approach can be adapted to other security challenges. Andrea Michi, the CTO of depthfirst, explained that owning the training process allows the team to optimize the model for what truly matters—detecting vulnerabilities efficiently. This customization makes the model cheaper to run, more effective at its tasks, and more adaptable to ongoing improvements.

The Growing Security Challenge in the Age of AI

Traditional security methods are struggling to keep pace with the rapid growth of software development. Companies are releasing more code, new services, and infrastructure updates faster than ever. Meanwhile, cyber attackers are also leveraging AI to find vulnerabilities more quickly and operate at a scale that older security tools can’t handle. This creates a widening gap between security needs and the capabilities of legacy systems.

depthfirst’s platform aims to bridge this gap by using AI that understands entire software systems. It can identify real risks and suggest specific fixes within developer workflows. The goal is to help teams secure their software as quickly as they build it, rather than treating security as a slow, separate process. This approach could revolutionize how security is integrated into software development, making it faster and more reliable.

The company emphasizes that AI will disrupt traditional security stacks. To succeed, security models need to be tailored for actual security workflows, using specialized data and domain-specific evaluation. depthfirst believes that building these models requires deep expertise and a focus on post-training development. Their team is among the few with this specialized know-how, positioning them well to lead innovation in AI-powered security tools.

Inspired by

• https://justainews.com/companies/funding-news/depthfirst-closes-80m-to-expand-ai-security-platform/