Skip to content

Sweet Security

Israeli startup providing runtime-based cloud detection and response (CADR) plus AI agent security, layered on top of traditional CSPM.

Visit Website ↗
65/100Incremental Innovator

Overview

Sweet Security builds a runtime-first Cloud Native Application Protection Platform that deploys a lightweight, eBPF-style sensor into cloud workloads to observe live process, network, and identity behavior. The platform combines Cloud Application Detection & Response (CADR), CSPM, identity threat detection, vulnerability management, and API security in a single runtime data plane, with a stated deployment time of around two hours.

In 2024 the company added generative-AI-powered detection and guardrails aimed at securing autonomous AI agents operating in cloud environments. Its customer logos include Fireblocks, Hippo, Kaltura, Cast & Crew, ShipStation, and SafeBreach.

Innovation Matrix Assessment

Innovation Velocity 7/10

Shipped a GenAI detection expansion in mid-2024 and AI-agent guardrails since, on top of its original 2023 CADR launch.

Operational Value 7/10

Runtime visibility plus a stated ~2-hour deployment window addresses real CNAPP alert-fatigue and time-to-value problems, though claimed accuracy figures are vendor-reported.

Market Momentum 5/10

Confirmed only a $12M seed round via PR Newswire; a subsequently reported Series A could not be independently re-verified, and no large enterprise logos or analyst placements were confirmed.

Category Disruption 7/10

Runtime-first CADR is a genuine architectural departure from posture-only CSPM/CWPP tooling.

Real-World Efficacy 5/10

No named breach prevention, independent benchmark, or third-party evaluation was found; efficacy claims are entirely vendor-sourced.

Enduring Relevance 8/10

Runtime cloud detection and AI-agent security sit squarely in the highest-growth segment of cloud security.

Why CISOs Should Care

Runtime detection cuts through the noise of posture-only tools by flagging what is actually running and exploitable in production.

What Makes It Different

Instead of scanning configurations and images before deployment, Sweet instruments the live runtime to see actual process, network, and identity behavior as it happens.

The Matrix Verdict

65/100 — INCREMENTAL INNOVATOR

A technically credible early-stage runtime CNAPP with a real architectural differentiator, but market and efficacy signals remain thin and largely vendor-sourced.

Editorial Note: Claims vs. Verified Findings

The $12M seed round is independently confirmed. A widely reported Series A and all performance/accuracy statistics could not be independently verified this session.

Sources