Rapid7, the Boston-based cybersecurity company listed on Nasdaq, said on March 26, 2026 that it had acquired Kenzo Security, a San Francisco startup that uses groups of AI agents to investigate security alerts without waiting for a human analyst.
The companies did not disclose the price or other terms. Rapid7 said it did not expect the deal to have a material effect on its revenue, annual recurring revenue, profitability or free cash flow.
Rapid7 said it would combine Kenzo's technology with the Rapid7 Command Platform and embed Kenzo's AI models across its managed detection and response (MDR) operations. Kenzo's co-founders, chief executive Harish Singh and chief product officer Partha Naidu, both commented on the deal in Rapid7's announcement.
What Kenzo built
Kenzo pulls security data from different tools into what it calls an entity-centric data mesh, normalising it so that software agents can analyse it together. Each agent specialises in a different security operations task, and together they investigate alerts, tune detection rules and run playbooks that change depending on what they find. Rapid7 said the platform supports all the functions of a security operations centre, including deploying detections and threat hunting.
Rapid7 cited a customer case study from Truework, in which the average customer saw investigation time fall by 94% and the share of alerts investigated rise from 12% to 100%. Naidu said in the release that in customer deployments investigation time had dropped from hours to minutes, and that the platform could investigate every alert instead of a sample.
When Kenzo came out of stealth in April 2025, Singh told SiliconANGLE that most AI security tools wrapped large language models around first-tier alert handling, and that Kenzo had built a platform rather than a chatbot.
“Kenzo Security advances our mission to enable security operations that preempt attackers with predictive precision and machine-scale.”
Why Rapid7 is buying
Rapid7 describes itself as a provider of AI-powered managed cybersecurity operations that combines exposure management with detection and response, and says it has more than 11,500 customers. Its announcement argued that most security teams can only investigate a fraction of the alerts their systems produce, leaving the rest unexamined.
Chief executive Corey Thomas said reactive security models had reached their limits, that attackers were using AI to move faster and that security teams were being asked to do more without extra resources. Rapid7 said that adding Kenzo would move its MDR service from AI-assisted workflows to AI-driven operations, with alerts investigated and correlated automatically across endpoint, identity, cloud, SaaS and third-party data.
Rapid7 also said the combined system would keep AI decisions auditable and let analysts decide where human sign-off is required, while agents continuously identify and prioritise exposures across customers' environments.
The founders and the funding
Singh and Naidu founded Kenzo in 2024, according to SiliconANGLE. Singh was previously a founder and founding engineer who worked at Lacework and E8 Security. Naidu is a former US Air Force cyber operations leader who later led security product development at Datadog and CrowdStrike.
The company announced a $4.5 million seed round in April 2025 from The General Partnership, a San Francisco venture firm, and Michael Coates, the former chief information security officer of Mozilla and Twitter. Built In SF reported at the time that Kenzo had spent 18 months in stealth and planned to grow to 20 people, focused on engineering and sales. Neither company gave Kenzo's headcount at the time of the acquisition.
What happens next
Rapid7 did not give a timeline for folding Kenzo into its products or say whether Kenzo would continue to sell its platform on its own. In its forward-looking statements, the company listed the usual deal risks, including the cost of integrating Kenzo's business and the possibility of failing to keep key staff.
Singh said Kenzo's models would keep improving outcomes as they ran, and that Rapid7's customer base would help make the technology more accurate.
