Doubling Down on Dropzone: 3 Learnings From Agentic Security Automation
Since we partnered with Edward Wu and Dropzone 18 months ago to lead their Series A, the company has established itself as the category leader in agentic security operation center (SOC) automation – actively protecting some of the world’s largest enterprises and MSSPs.
As their growth accelerates, we’re excited to lead the company’s Series B, joined by our friends at Madrona, along with existing partners Decibel Partners, Pioneer Square Ventures, and In-Q-Tel.
The space has evolved rapidly in the past year: here are three of our key learnings on building AI native products in complex domains like security:
1. Enterprises are ready to use AI in mission-critical workflows
Last year, most enterprises still viewed AI with caution. While employees used AI assistants, most execs were still skeptical that it could reliably automate mission-critical workflows, particularly in sensitive areas like security.
Now, mission-critical AI automation has moved into the mainstream, with broad excitement and adoption in security.
Three things have catalyzed this change:
- AI models and systems have improved >10x, reliably (and cost-effectively) handling many complex enterprise tasks.
- Security and risk teams have become comfortable with the deployment model. When an enterprise already trusts hyperscalers with their data, it’s reasonable they can trust those same companies running LLMs.
- Leading companies have adopted AI systems like Dropzone, demonstrating the impact and decreasing the perceived risk of failure.
For Dropzone, this has resulted in dramatic acceleration in the sales pipeline over recent quarters, and production deployments at larger enterprises including Fortune 500 companies and top-10 MSSPs.
2. AI security products sell outcomes, not insurance
Historically, most cybersecurity products have sold insurance-type products: they protect against future potential risks, and “working well” means “nothing happens”. These products are hard to evaluate, so are sold based mostly on relationships and reputation. This advantages sales-led organizations and executives with close personal relationships to CISOs.
Emerging AI security products are different. While they dramatically improve security posture and outcomes, they do so by automating human work. That makes them much easier to evaluate: How many alerts did they resolve? How quickly? How accurate were their responses compared to humans?
With a clear scoreboard, companies can no longer coast on overambitious marketing and backchannel introductions. Sales, marketing, and channel are still essential, but only work when there is an exceptional product to sell. Dropzone has won nearly 100% of its technical evaluations against both startups and incumbents, giving us confidence that their product speaks for itself at any discerning enterprise.
3. Effective automation requires AI-native products and AI-native teams
As the value of AI SOC analysts became clear, more companies have pivoted or rebranded themselves as agentic AI – from SIEMs, to previous gen SOAR, to security copilots. But agentic systems are an entirely new kind of software to build.
They need:
- Thousands of prompts, tools, inference calls, and adaptation mechanisms built with deep domain knowledge to make automation work reliably.
- New infrastructure to support data processing, retrieval, and system orchestration.
- Novel interfaces and UX flows for humans to collaborate with AI.
It’s effectively a wholesale change from existing software, which means it’s much easier to start from scratch than by pivoting a legacy product.
Even strong software development teams struggle with agent-native products. They are complex systems, and work is non-deterministic both for each component and the overall system level. For teams that are used to building predictable, deterministic software, it makes development a nightmare. Dropzone’s DNA is AI-native and experimentation-oriented at its core, all the way up to the CEO Edward. Their product development has been focused and fast, and as they expand out from their initial functionality (more to come soon) we feel confident they can continue to out-execute the competition.
—
We are confident that Dropzone will be the leader in AI-native security operations, and thrilled to continue our partnership with the team. To learn more about the platform, request a demo here.
