Terms of Service

Access is granted after review, not self-served. We may approve, decline, suspend, or revoke access at our discretion, including for suspected abuse, security risk, or breach of these terms.

Access is tied to your tenant workspace and the role you are assigned within it. We may change or remove entitlements as an engagement evolves.

Sign-in uses an email magic link plus a TOTP two-factor code. Keep your sign-in email account secure, since anyone who controls it can request a link, and keep your second factor private.

You are responsible for activity that occurs under your account. Notify us promptly at security@spaltx.com if you believe your account or email has been compromised.

Do not probe for vulnerabilities without prior written authorization, and do not attempt to reach, read, or infer another tenant's data.

Do not use the AI agent tooling to exfiltrate data or to reach internal services, private network ranges, or cloud metadata endpoints, and do not attempt to bypass tool allowlists or other agent controls.

Do not upload content you lack the rights to process, malware, or unlawful material. Our full Acceptable Use Policy governs these obligations and forms part of these terms.

We aim for high availability but provide the platform on an as-available basis during this build phase, without a contractual uptime commitment.

Invoices shown in the portal are for display and status only. They reflect the state of billing for your engagement. No payment is processed through the platform in v1; settlement happens through the channels agreed in your engagement.

You own your tenant content. We process it only to provide the service, under the isolation, encryption, and access controls described in our Privacy Policy and Security pages.

We do not use your tenant content to train models. Each party will protect the other's confidential information with reasonable care.

AI and agent output can be imperfect, incomplete, or wrong. You are responsible for reviewing outputs before relying on them, especially for decisions with legal, financial, or compliance consequences.

The document-extraction pipeline keeps verbatim capture separate from decisions and routes low-confidence output to human review. These are quality controls, not guarantees of correctness; the duty to verify before relying remains with you.

To the extent permitted by law, the platform and any AI output are provided without warranties of any kind, express or implied, including fitness for a particular purpose.

To the extent permitted by law, our aggregate liability arising out of or relating to the platform is limited to the fees paid for the applicable engagement.

We may update these terms as the platform evolves. Material changes will be reflected in the “Updated” date above, and continued use after a change constitutes acceptance.

Governing law and venue will be specified in your engagement agreement.