Service Terms Service Terms

1. Purpose

These terms govern the use of Kujira, the artificial intelligence agent orchestration platform operated by Calltek Network SL ("we", "Kujira"). By creating an account or using the service, you agree to be bound by them.

If you register or act on behalf of a company or another organization, you represent that you have sufficient authority to bind it, and "you" or "your organization" refer to it throughout this document.

2. Definitions

So we don't have to repeat the explanation every time they come up, these terms mean:

• Platform / the service — the web application, API, and panel that we operate, taken together.
• Control plane — the part we host: accounts, orchestration, billing, panel. It does not include the server where your agents run.
• Server — the equipment you provide (a VPS, an office computer, a NAS) where each agent physically runs.
• Agent — an AI process configured within your organization, with its own role, credentials, and communication channels.
• Organization — the workspace under which your agents, servers, and team members are grouped.
• Content — any data, file, credential, or conversation that you or your agents enter or generate while using the service.
• Professional services — the support contracted separately under the Enterprise plan: consulting, initial setup, training, and follow-up.

3. Capacity to contract

You must be of legal age and have the legal capacity to enter into a contract. If you register on behalf of a company, you are responsible for having the authority to do so and for the accuracy of the information you provide.

The service is primarily intended for professional or business use. If you register as an individual and use it outside any professional activity, you act as a consumer, and certain clauses in this document —expressly marked, see clause 16— protect you under Spanish law regardless of what the rest of the text says.

4. How it works

Kujira deliberately keeps two parts separate. We operate the control plane: the account, the panel, the orchestration of your agents, and billing. The agents themselves run on your own server and talk to the artificial intelligence model that you choose and contract on your own —Anthropic or another provider—.

That separation has a direct consequence for everything that follows: we do not host your agents' conversations in the sense of "owning" them, and we are not responsible for the availability of your server or of the model provider you chose. What we do is store, index, and give you access to that information through the panel, with the retention period you configure.

5. Account and organization

Each organization has one person with the OWNER role, responsible for who else gets in and with what permissions. Roles are not cosmetic: they determine who can create agents, manage credentials, invite people, or delete the organization, and that assignment is your responsibility, not ours.

Your password and, if you enable it, two-factor verification are your responsibility: if someone logs in with your credentials, they act as you for all purposes in our relationship with you. Notify us as soon as you suspect they have been leaked or that someone is accessing your account without authorization.

6. Pricing and billing

You pay per active agent, monthly, with no base fee and no minimum term. The exact amounts —with and without VAT— are on the pricing page, which is the source that governs: if anything here conflicts with it, pricing prevails.

Billing is handled through our payment gateway; your card details never pass through our servers. The invoice is issued at the frequency of your contracted plan and reflects the agents you had active that month.

7. Duration and cancellation

The contract runs for an indefinite term and you can cancel it whenever you want, from the panel itself. There is no penalty or minimum term.

Stopping payment or canceling doesn't delete anything right away: the organization drops to the FREE plan, your agents, memories, and conversations stay where they were, and only one keeps running until you decide otherwise. If you are a consumer and want to exercise the 14-day right of withdrawal under distance-selling regulations, keep in mind that starting to use the service immediately —creating an agent, for example— may mean, if you expressly accept it when contracting, the loss of that right for the part already performed.

8. Acceptable use

Agents act with the access you give them: vault credentials, email, repositories, external channels. That access is your responsibility, just as it would be if you gave it to an employee. By using the service, you agree not to use it, or your agents, for:

• Engaging in illegal or fraudulent activities, or activities that infringe third-party rights —intellectual property, data protection, competition law—.
• Impersonating a person or entity, or deceptively passing off automated communications as human.
• Sending spam or unsolicited mass communications through connected channels.
• Breaching, scanning, or attempting to access without authorization systems that are not yours, regardless of whether it is an agent doing it and not you directly.
• Circumventing the service's technical limits —on agents, channels, or usage— or reselling access to third parties without an express written agreement.
• Reverse-engineering the panel or the orchestration, except to the extent mandatorily permitted by law.
• Breaching the terms of use of the third parties you connect an agent to —the model provider, GitHub, Google, and similar—, which is your responsibility to comply with, not ours.

9. Agent risks

An agent is a language model that executes instructions automatically. It can make mistakes, misinterpret a request, or produce incorrect, incomplete, or unexpected results —what is known as a "hallucination"—. We do not guarantee the accuracy, completeness, or suitability of anything an agent says or does.

You decide the scope: what credentials you give it, what channels it manages, how much spending you authorize, and which actions require your approval before they run (the human approvals system exists precisely for that). The more critical or irreversible an action is —sending a payment, deleting something, contacting one of your clients—, the more you should require human review before the agent carries it out. Choosing not to delegate that review is your decision, not ours.

10. Third-party access

When you connect a credential, an email account, a repository, or a calendar to an agent, you are giving it real access to that system. We encrypt those credentials at rest, control who in your organization can view or operate them, and keep an audit log of every access —who revealed what, and when—.

What we don't do is decide the scope of that access for you, or take responsibility for what a third party does with the data once your agent sends it to them —your email provider, the repository, the model provider itself—. Granting broader access than necessary is a risk you take on when you configure it.

11. Intellectual property

The software, the design, the Kujira brand, and everything that makes up the control plane are our property or are used by us under license. For as long as your contract lasts, we grant you a non-exclusive, non-transferable license limited to accessing the service under these terms.

The content you enter, or that your agents generate —conversations, documents, configurations— remains yours. We do not use it to train artificial intelligence models, and we do not share it with anyone beyond what is already explained in the privacy policy.

12. Privacy

What data we process, for what purpose, and for how long is explained in the privacy policy. If your organization processes your own clients' data through your agents, the split of responsibilities between you and us as data processor is set out in the data processing agreement.

Retention of conversations is configurable per organization: you decide how long they are kept. Every sensitive action —creating an agent, revealing a credential, changing someone's role— is recorded in your organization's audit log.

13. Confidentiality

Each party agrees to protect the other's confidential information that it becomes aware of because of the contract —configuration, business data, credentials— with the same care it applies to its own, and not to use it for purposes other than providing or receiving the service.

14. Availability

We take care to keep the control plane up, but this is not a service level agreement with financial penalties: if you need that commitment in writing, that is what gets negotiated under Enterprise.

Some features are expressly marked as beta: they may change or fail more often than the rest, and their availability is not covered by any additional warranty. Whoever operates your own server, and the model provider you chose, is responsible for its availability, not us.

15. Warranties

We provide the service "as is" and "as available". To the extent the law allows, we do not warrant that it will be uninterrupted, error-free, or fit for any particular purpose other than what this website describes, unless we have expressly agreed otherwise in writing —typically in an Enterprise agreement—.

16. Liability

To the extent the law allows, our total liability to you for any claim arising from this contract is limited to the amount you have paid for the service in the 12 months prior to the event giving rise to it. We are not liable for indirect damages, loss of profit, loss of business, or data loss resulting from your own configuration —the scope you gave an agent, for example—.

This limit does not apply to what the law does not allow limiting: willful misconduct, gross negligence, or personal injury. If you are contracting as a consumer, this limit applies without prejudice to the rights that consumer protection law grants you on a mandatory basis.

Nor are we liable for the acts, omissions, or unavailability of third parties that you choose and contract with —the artificial intelligence model provider, your own server, the services you connect through a channel or connector—, or for decisions an agent makes within the scope you configured for it.

17. Indemnification

If you misuse the service, breach these terms, or the scope you gave an agent causes harm to a third party, you agree to hold us harmless against any claims arising from it, including reasonable defense costs.

18. Suspension and termination

We may suspend or close an account if we detect use that breaches clause 8 —abuse, illegal activity, a security risk to other clients— or sustained non-payment, giving advance notice unless the urgency of the risk prevents it.

When you close your account, your data is deleted according to the timeframes in the privacy policy, except for what we must keep under a legal obligation.

19. Changes

The service evolves: we may add, change, or remove features, giving reasonable advance notice when the change is significant to your use.

If we substantially change these terms, we will notify you by email or from the panel itself before they take effect. Continuing to use the service after that date means accepting them; if you disagree, you can cancel under clause 7.

20. Force majeure

Neither party is liable for a failure to perform caused by circumstances beyond its reasonable control —including outages at infrastructure providers or third-party model providers— for as long as those circumstances last.

21. Assignment

You may not assign this contract to a third party without our consent. We may assign it as part of a corporate transaction —a merger, the sale of the business— by notifying you and without that reducing your rights as a client.

22. Communications

Legal communications relevant to you are sent to your account email or through the panel itself. For any question, write to us at soporte@kujira.so.

23. General provisions

If any clause of this document is found void or unenforceable, the rest remains in effect. Our not strictly enforcing something on a given occasion does not mean we waive the right to enforce it later. This document, together with the pricing page and the policies it links to, is the entire agreement between the parties regarding its subject matter.

24. Governing law

These terms are governed by Spanish law. For any dispute, the parties submit to the courts that have jurisdiction under applicable law —which, if you are contracting as a consumer, are the courts of your own domicile—.