cinchy.com

What It Is

Cinchy's PeriMind is an enterprise AI Governance Platform positioned as a registry and runtime control plane for AI tool calls. It is not a large language model itself; instead, it sits between AI interfaces and enterprise tool endpoints, governing how ChatGPT, Claude, Gemini, in-house Agents, IDE Copilots, and similar systems call MCP Servers, Skills, APIs, CLIs, databases, and cloud infrastructure.

Core Capabilities

The platform emphasizes that “every tool call is governed”: it first verifies the Agent’s identity, then handles authorization, rate limiting, OPA/Rego policy enforcement, payload inspection, and audit logging. Key capabilities include a registry for tool endpoints, scoped credentials per Agent, fine-grained policies, Agent/tool-level quotas, hash-chain tamper-resistant audit trails, chain-of-thought capture, and a federated governance model. Compared with a conventional API Gateway, PeriMind’s differentiation is that it attempts to understand the semantics, intent, and compliance context of tool calls, rather than merely forwarding traffic.

Pricing and Trial

The website does not disclose public pricing, plans, or a free tier. It only provides Get Started, Request a Demo, and Book a Call options. Its legal terms indicate an enterprise licensing model, with fees determined by an Order Form/SOW. Pricing is typically based on purchased Licenses rather than actual usage; fees are prepaid or invoiced according to the order, payable within 30 days, and payment obligations are non-cancelable and non-refundable.

Pros and Cons

The main advantage is that it targets a very relevant use case: AI Agents are already beginning to call enterprise systems directly, while traditional IAM, DLP, SIEM, and WAF tools struggle to understand the semantics of tool calls. PeriMind covers discovery, registration, runtime control, auditing, and compliance evidence, making it suitable for enterprises with high security and compliance requirements. The limitation is that the public information still feels more like product marketing: there is a lack of disclosed performance data, deployment models, real-world case studies, and measurable outcomes. Chinese-language support, data residency, subprocessors, and specific security certifications are also not sufficiently detailed in the main content.

Who It’s For and Access from China

PeriMind is better suited to medium and large enterprises, security teams, data governance teams, and platform engineering teams that already have extensive Copilot/Agent/API/MCP integrations. It does not appear to be a self-service tool for individuals or small teams. The official content does not state how well it can be accessed from China, and payment methods can only be inferred as enterprise invoice-based payments. For deployment in China, buyers should carefully evaluate network connectivity, contract and payment arrangements, cross-border data transfer, Chinese-language support, and potential alternatives such as enterprise API gateways, AI security gateways, identity governance tools, DLP/SIEM platforms, or localized AI governance solutions.