Terms of Service

These terms govern access to and use of the AI Agent Inference product ("the Service") provided by BastionShield Technologies Ltd ("BastionShield", "we", "us"). They apply alongside any separate written commercial agreement between us; where that agreement and these terms conflict, the commercial agreement prevails.

The Service is offered on a sales-led basis. There is no self-service sign-up; access is arranged through a commercial engagement.

The Service is an inference gateway that sits between your applications or agents and the model providers you configure. It carries inference requests to a configured model endpoint, handles responses and errors, and records operational metadata such as usage and latency.

The Service is infrastructure for running inference. It is not a governance or policy-enforcement product, and it does not decide what your agents are permitted to do.

Product interfaces, dashboard figures, sample metrics and code examples shown on our website and in materials are illustrative. They are provided to explain how the Service works and do not represent guaranteed performance, availability, throughput, latency, cost or production results. Actual behaviour depends on your configuration, the model providers you use and your deployment.

You configure which model providers and models the Service routes to. Your use of those providers is subject to their own terms, pricing and acceptable-use policies, and you are responsible for holding the necessary accounts, credentials and rights.

Provider availability, changes, deprecations and outages are outside our control. Where reliability features such as timeouts, retries or fallback are configured for your routes, they operate within the limits of what your providers make possible.

You agree not to use the Service unlawfully, to infringe others’ rights, to attempt to disrupt or reverse-engineer it, or to route content you are not permitted to process. You are responsible for the requests your applications and agents send through the Service.

The Service is designed so that the content of requests and responses is not retained; we process operational metadata to run, secure and meter it. Our handling of data for the Service is described in the AI Agent Inference Privacy Policy, which forms part of these terms.

Unless a separate written agreement states specific service levels, the Service is provided without a committed availability level, and support is provided as described in your commercial engagement. We may modify, suspend or discontinue parts of the Service, and will give reasonable notice of material changes where we can.

To the extent permitted by law, the Service is provided "as is" and "as available". We do not warrant that it will be uninterrupted or error-free, or that it will meet requirements not set out in a written agreement. We do not claim compliance with any framework or certification we do not hold.

Nothing in these terms limits liability that cannot be limited by law. Subject to that, we are not liable for indirect or consequential loss, or for loss arising from your model providers, your configuration, or content you route through the Service. Any aggregate liability limit is as set out in your commercial agreement.

These terms apply while you use the Service. Access may be ended in line with your commercial agreement, or by us where use breaches these terms. Provisions that by their nature should survive termination will do so.

These terms are governed by the laws of England and Wales, and the courts of England and Wales have exclusive jurisdiction, without affecting any mandatory rights you have under local law.

Questions about these terms: info@bastionshieldtechnologies.com, or BastionShield Technologies Ltd, 124 City Road, London, EC1V 2NX, United Kingdom.