Cygrid
Cygrid is an identity enforcement platform for AI-driven enterprises, mapping human, non-human, and AI agent identities into a single graph to control what each can access and do. The platform enforces policy at three points—pre-use, out-of-band, and in-band at runtime via an MCP Gateway—blocking or rewriting risky agent actions in milliseconds. It connects read-only to identity providers, SaaS, cloud, CI/CD, and AI layers through native APIs, ingesting existing security tools as feeds rather than replacing them.
- Cybersecurity
- Artificial Intelligence
- Enterprise Software
- Software Only
Funding
Raised to date
€2.7MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Est. $2.9Macross 1 round
€2.7M · Pre-Seed
Announced Oct 2026
Disclosed cumulative: $2.9M
Founders
Founder details are not available yet.
Product
Problem
Most cyberattacks start with a stolen identity, and AI agents now act on behalf of people with real access to business systems. Every new agent is another identity to compromise, leaving companies stuck between blocking AI entirely or accepting identity risk they cannot see or stop.
Solution
Cygrid is the identity enforcement platform for AI-driven enterprises. It maps human, non-human, and AI agent identities into one Identity Intelligence Graph, scores what each identity can do, and enforces policy wherever an identity can act. The platform connects read-only to identity providers, SaaS, cloud, CI/CD, and the AI layer through each system's own APIs, ingesting existing security tools as feeds rather than replacing them. Enforcement happens at three points: pre-use, where new agents get scoped, short-lived credentials; out-of-band, where drift from policy is corrected at the source in seconds; and in-band, where agent tool calls are blocked or rewritten at runtime per parameter at the MCP Gateway. Every verdict is written to an immutable decision log.
Target Audience
Primary customers are AI teams, IT operations, and security and compliance functions within AI-driven enterprises, particularly in financial services and e-commerce, that need to use AI safely while enforcing identity policies across human and non-human actors.
Features
- Identity Intelligence Graph unifying humans, non-human identities, AI agents, tokens, and MCP servers with risk scoring from policy rules and AI-driven analysis
- MCP Gateway for in-band runtime enforcement that blocks or rewrites agent tool calls per parameter in milliseconds
- Pre-use enforcement that checks every new agent, app, or integration against the graph and issues scoped, short-lived credentials instead of standing grants
- Out-of-band remediation that corrects identity, permission, or configuration drift at the source in seconds without waiting for review cycles
- Read-only source connections via native APIs to identity providers, SaaS, cloud, CI/CD, and AI layers, with existing security tools ingested as feeds
- Immutable decision log recording every allow, downscope, block, revoke, or just-in-time grant verdict