Maze, a cybersecurity startup developing autonomous AI agents to investigate and resolve cloud vulnerabilities, has raised $25 million in Series A funding. The round was led by Theory Ventures, with participation from Cherry Ventures and Tapestry VC.

Founded by Harry Wetherald, Maze is building next-gen tooling for DevSecOps teams who are overwhelmed by the volume, complexity, and urgency of modern cloud security threats. The company’s mission is simple but bold: give security engineers an intelligent system that doesn’t just detect vulnerabilities  -  but fixes them autonomously.

As cloud environments grow more complex and elastic, security tools that only offer alerts are no longer enough. Maze’s autonomous agents aim to close the gap between detection and remediation  -  in real time.

---

What Maze Is Building

Maze is not just another alerting tool or dashboard. It’s building a network of AI security agents that act like an internal security team - 24/7, tireless, and precise.

Key capabilities include:

• Autonomous vulnerability triage: AI agents scan cloud environments, prioritize threats based on business impact, and assign risk levels.
  
• Real-time remediation: Agents can resolve misconfigurations, permissions issues, and known exploits autonomously using pre-approved playbooks.
  
• Security observability layer: Provides full traceability into what the agents saw, why they acted, and how the issue was resolved.
  
• Cross-cloud compatibility: Works across AWS, Azure, and GCP, including containerized and serverless environments.
  
• Team integrations: Slack, Jira, and GitHub hooks to keep human teams in the loop without overwhelming them with false positives.
  

Think of Maze as a self-driving car for cloud security  -  instead of telling you that a crash is imminent, it grabs the wheel and steers you to safety.

---

Why It Matters

Today’s DevOps and security teams are outgunned.

• The average enterprise uses over 130 SaaS applications and operates across multi-cloud environments, creating a massive attack surface.
  
• 60% of cloud breaches are caused by misconfigurations or human error (Gartner, 2023).
  
• The mean time to remediate a cloud vulnerability is still over 15 days, while attackers can exploit misconfigs in hours (IBM X-Force Threat Intelligence Index).
  
• Security teams are burned out by alert fatigue, dealing with tens of thousands of notifications that often lack prioritization or context.
  

Maze is built for this reality. It's not trying to replace security engineers - it’s trying to amplify them with intelligent, tireless, AI-powered agents.

---

A Strategic Insight for Founders

Amid the buzz of AI and automation, Maze delivers a deeper lesson: Build for the operational layer, not the observability layer.

Rather than analyzing faster, Maze executes faster. It closes loops instead of surfacing them.

That distinction is where most technical products miss. Founders often build tooling to make humans smarter. Maze built a system to make humans unnecessary at critical junctures  -  not by removing control, but by removing repetition.

Here’s the real unlock: Maze enables teams to define what “secure enough” looks like and lets agents enforce it automatically. That frees engineers to build without context-switching and lets security leaders focus on edge cases instead of firefighting.

Founders building in fast-moving technical domains should ask: Where can I safely automate the burden without asking users to change their behavior? That’s product-market fit at a systems level.

---

Market Outlook: Why Maze Is in the Right Place, Right Time

The cybersecurity and cloud automation industries are both surging:

• Cloud security spending is projected to exceed $97.3 billion by 2032, with enterprise demand outpacing available talent (Allied Market Research)
  
• The DevSecOps tools market is growing at 31% CAGR, fueled by the need to embed security into CI/CD workflows (MarketsandMarkets)
  
• Enterprises now deal with an average of 3,900 cloud misconfigurations per month, many never investigated (Orca Security, 2023)
  
• 75% of security professionals believe they would benefit from AI automation for threat detection and response (Splunk State of Security Report, 2024)
  

Maze is building at the intersection of all of these pressures  -  and solving not with more dashboards, but with invisible, intelligent action.

---

Who Maze Is Built For

Maze is designed for cloud-native engineering teams who:

• Operate across multi-cloud or hybrid cloud environments
  
• Use Kubernetes, serverless, or containerized workflows
  
• Have overextended security teams chasing down misconfigs
  
• Want to enforce continuous security without slowing down dev velocity
  
• Need traceable, auditable remediation with zero trust by design
  

From fintech to e-commerce to AI infrastructure startups, the shared pain is the same: complexity is growing faster than teams can hire for it.

Maze gives these teams a way to fight complexity with autonomy.

---

What’s Next for Maze?

With the fresh Series A funding, Maze plans to:

• Expand its AI agent library across new remediation categories
  
• Build custom policy frameworks for regulated industries like healthcare and finance
  
• Invest in explainability and audit tools, ensuring full transparency for every agent action
  
• Grow its go-to-market team to onboard DevSecOps teams globally
  
• Launch sandbox testing features so security leads can simulate agent actions before deployment
  

The long-term vision? To become the default autonomous cloud security layer in the modern DevOps stack - one that’s always-on, always-updating, and always acting in the best interest of uptime, safety, and compliance.