Philippe Suchet
Founder
Onyx
Onyx offers a unified security and governance platform that continuously monitors, controls, and audits AI agents across SaaS, cloud, endpoint, and code environments. Its real‑time Guardian Agent enforces natural‑language policies, blocks prompt injection and data exfiltration, and provides compliance reporting and cost‑optimization for enterprise AI deployments.
- Artificial Intelligence
- AI Agents
- Cybersecurity
- Enterprise Software
Updated 2 days ago
Funding
Funding not disclosed
Funding rounds are not available yet.
Founders
Shira Kaplan
Co‑Founder
Dirk Marich? ??
Co‑Founder ??
Product
Problem
Enterprises are rapidly deploying AI agents across SaaS, cloud, endpoint, and code environments, but lack visibility and control, leading to data exposure, excessive permissions, and vulnerability to attacks such as prompt injection and remote code execution.
Solution
Onyx provides a unified security and governance platform that continuously monitors and controls AI agents throughout an organization’s stack. A supervisory Guardian Agent inspects every prompt, response, and action in real time, detecting anomalies, shadow AI usage, and supply‑chain risks. The platform enforces policy rules—defined in natural language and aligned with standards like MITRE, NIST, and the EU AI Act—across all agents, models, and MCP servers. Integrated AI observability, security posture management, and orchestration tools give security, governance, and AI teams full audit trails, compliance reporting, and cost‑optimized deployment capabilities without disrupting workflows.
Target Audience
Primary customers are security, compliance, and AI operations teams in large enterprises that run AI agents across SaaS applications, cloud services, endpoints, and custom code.
Features
- Real‑time AI observability with session replay, audit trails, and multi‑cloud, multi‑agent unified view
- AI Security Posture Management that blocks prompt injection, jailbreaks, data exfiltration, and adversarial manipulation
- Natural‑language policy engine with pre‑populated templates for regulatory frameworks and industry standards
- Guardian Agent that autonomously identifies risks and remediates issues across thousands of agents
- Managed AI gateway that routes traffic, performs failover, load‑balancing, and cost optimization for LLM inference
- 100+ pre‑built integrations (AWS, Azure, GCP, OpenAI, Anthropic, CNAPP, SASE, EDR, SIEM/SOAR)
- Flexible deployment options: cloud, hybrid, or self‑hosted with support for on‑prem VPC and custom proxies
- Executive AI ROI dashboards that track adoption, productivity gains, and total AI spend