Terms of Service

The terms that apply to using the Workforce Intelligence connection service, Workforce Intelligence inference service and account portal, and documentation site.

Effective · Last updated

Acceptance and scope#

These terms are a contract between you and Workforce Intelligence, Inc. ("we", "us"). They cover the Workforce Intelligence connection service, the Workforce Intelligence inference service and account portal, and the public documentation site (together, the "Service"). By using the Service you agree to them. If you use it for an organization, you confirm you may bind that organization.

These terms do not cover a customer's own Workforce Intelligence installation. Its operator sets the terms that apply to it.

Who may use the Service#

The Service is for customers' Workforce Intelligence installations and the people those customers authorize to connect accounts through them or to manage the customer's Workforce Intelligence account. The documentation site is open to everyone. The Service is not for children under 16.

Your Workforce Intelligence account#

The Workforce Intelligence inference service (the programming interface at /v1) answers requests from a customer's installation with responses from AI models. The account portal (the website at /portal) is where the customer manages the account that pays for them.

  • To create an account you give an email address you can receive mail at and a password of at least 12 characters, and confirm the address with the single-use link we send. The link works for 24 hours.
  • Each email address can have one account, and each sign-in belongs to one account.
  • You keep your password secure. You are responsible for what is done while signed in to your account.
  • A Workforce Intelligence account is for the customer's own installations. Each of the customer's installations uses an API key from the customer's account.

API keys#

You create API keys in the portal. A key is shown once, when you create it, and we cannot show it again. You keep each key secret and are responsible for every request made with it, including its charges. You can have up to 20 active keys at a time. To replace a key, create or rotate it, put the new key in your installation and revoke the old one; a revoked key stops working within about a minute. If a key may have been exposed, revoke it at once and write to privacy@getwi.com. Use each key only for the customer's own installations.

Charges and payment#

  • What is charged. Use of the inference service is charged by use, measured in tokens (or in images, for image generation) for the model that answered, at our rates. We set the rates and can change them, and a change applies to requests made after it takes effect. The usage statement in the portal lists each request and is the basis for your invoice.
  • When and how. The billing month is the calendar month in UTC. After a month ends we issue one invoice for it, in US dollars, and Stripe charges it automatically to the card on file. You add the card through a page hosted by Stripe. Without a card on file, requests are refused.
  • Rounding. Invoices are in whole cents and are never rounded up. Any part of a cent, and any month's total below US$0.50, is added to the next month's invoice.
  • Failed payment. If a payment fails, Stripe may try again. If an invoice is finally marked uncollectible, we suspend the account and requests are refused until the invoice is paid, and we restore access when it is paid unless the account is suspended for another reason.
  • Questions about an invoice. Write to privacy@getwi.com.
  • Accounts we provide at no charge are not invoiced and are subject to a monthly spend limit that we set.

Spend limits and request limits#

  • Spend limit. You can set a monthly spend limit for your account in the portal, in US dollars. It applies from your next request and starts again at the beginning of each month (UTC). Without a limit there is no spend limit.
  • Pacing. With a limit set, spending is spread across the month. If you use your budget faster than an even pace, some requests are refused with a "try again after" response, and your installation waits and retries for a limited time, so those requests take longer.
  • Stopping at the limit. Once the spend recorded for the month has reached the limit, new requests are refused until the next month or until you raise the limit. A request that starts while you are under the limit is always completed and its full cost is charged, and requests started at the same moment can each be accepted, so one long request, or several at once, can take the month's total above the limit by their whole cost.
  • Request limits. We also limit how fast requests can be sent, for each API key and for each source address. A request over a limit is refused, and when the service is at capacity it asks you to try again shortly. Requests can be delayed or refused.

Models and availability of responses#

  • Names. Workforce Intelligence's own model names (reasoning, vision, embedding and image, each at a low, mid or high level) stand for a level of quality and cost. We choose which model answers a request, it can differ between requests, and it can be from a different provider or model family. Model family names (for example claude-haiku, claude-sonnet, claude-opus, gpt-luna, gpt-terra, gpt-sol, gemini-flash-lite, gemini-flash, gemini-argon and grok) use the newest version of that family we offer. If the provider of that model cannot answer, a request can use the same model through Amazon Bedrock where Bedrock offers it, or an equivalent model of the same level (low, mid or high) from a different provider. Charges are based on the model that answers a request and the provider that serves it: when the same model is served through Amazon Bedrock, the request is charged at Bedrock's rate for that model, which can be higher. When a request is answered by an equivalent model from a different provider, the price of that request follows the model that answered it, so it can be higher or lower than it would have been, and your usage statement marks such requests as answered by an equivalent model.
  • No guarantee. We do not guarantee that a model, a model version or a response is available, or that responses are accurate, complete, or suitable for your purpose. We can add, change or remove models and names. You are responsible for checking and using responses appropriately.
  • Providers. Requests are answered by third-party AI model providers, whose terms and policies apply to what they do with the content they receive. The Privacy Policy names them and describes what we send them.
  • Your content. You are responsible for the content you send and for having the right to send it. You keep all rights you have in that content and in the responses you receive. We process request content only to produce the response and do not store it.

What the Service does#

The connection service lets an installation connect to accounts at the outside services it is set up to use, without registering its own application with each one. When a person starts a connection, we complete the provider's sign-in step and give the result to the installation that asked for it, using a sealed record that is accepted for at most 5 minutes and is deleted when first collected. When an installation asks, we also refresh a connection's tokens. We do not store tokens, and we do not read the content of connected accounts. The Privacy Policy and Data handling and security describe this in detail, together with how the inference service and account portal handle data.

Acceptable use#

You will not:

  • abuse, probe or test the security of the Service without our written permission;
  • circumvent rate limits, access controls or other protections;
  • overload the Service or send traffic meant to degrade it;
  • resell, sublicense or provide access to the connection service, the inference service or a Workforce Intelligence account to anyone other than the customer whose installation uses it, or use an API key anywhere other than in that customer's own installations;
  • share or publish an API key, or let another person sign in as you;
  • use the inference service to break the law, or the terms and usage policies of the model providers that answer your requests;
  • use the Service to break the law or the terms of a provider whose account you connect;
  • interfere with other people's connections.

The customer's responsibilities#

The customer is responsible for its installation, the people it authorizes, the accounts those people connect and what the installation does with the tokens it receives. The customer keeps the access and refresh tokens it receives secure, and uses them in line with each provider's terms and policies. The customer is also responsible for its Workforce Intelligence account: its password, its API keys, the requests made with them, the content of those requests and the charges that result.

Third-party services#

The outside services that an installation connects through the Service, the AI model providers that answer requests, and Stripe are independent of us. Their own terms and privacy policies apply to your accounts and data with them, and we are not responsible for their services, their decisions to grant, limit or revoke access, or their changes.

Availability and changes#

We run the Service with care, but we do not guarantee any level of uptime or availability unless a separate written agreement with us says otherwise. We can change, limit or discontinue any part of the Service. A Workforce Intelligence installation does not depend on the connection service to run: when it is unavailable, the installation continues to work and offers its own alternative way to connect accounts. Requests an installation sends to the inference service fail while it is unavailable; models the installation reaches with its own provider keys are not affected. The inference service and the account portal are available from the United States, Canada and the member states of the European Union, and requests from other countries are blocked.

Rights in the Service and documentation#

We and our licensors own the Service and the documentation. We grant you a limited, non-exclusive, non-transferable right to use the Service as these terms allow and to read and share the documentation for your own use of Workforce Intelligence. You keep all rights in your accounts and data. No other rights are granted.

Disclaimers#

The Service is provided "as is" and "as available". To the fullest extent the law allows, we disclaim all warranties, express or implied, including merchantability, fitness for a particular purpose and non-infringement. We do not warrant that the Service will be uninterrupted or error-free.

Limitation of liability#

To the fullest extent the law allows, we are not liable for indirect, incidental, special, consequential or punitive damages, or for lost profits, revenue, data or goodwill, arising from the Service. Our total liability for any claim arising from the Service is limited to the greater of the amount you paid us for the Service in the twelve months before the claim and one hundred US dollars. Nothing in these terms limits liability that cannot be limited by law.

Indemnity#

Each party will cover the other for third-party claims, and resulting losses and reasonable costs, that arise from the first party's breach of these terms or violation of law. The party seeking cover gives prompt notice, allows the other to control the defense and cooperates reasonably.

Suspension and termination#

You can stop using the Service at any time; to close a Workforce Intelligence account or have its data deleted, write to privacy@getwi.com. Charges for use before you stop remain payable. We can suspend or end your access, or an installation's access, or a Workforce Intelligence account or API key, at once if you breach these terms, if your use threatens the Service or others, if an invoice is uncollectible, or if the law requires it. Sections that by their nature should continue, including rights in the Service, disclaimers, limits on liability, indemnity and governing law, continue after termination.

Changes to these terms#

We can update these terms. We will change the "Last updated" date, and for a material change we will publish it before it takes effect and change the "Effective" date. Using the Service after a change takes effect means you accept it.

Governing law and venue#

These terms are governed by the laws of the State of Delaware, without regard to its conflict-of-laws rules. Disputes are brought in the state and federal courts located in the State of Delaware.

Contact#

Workforce Intelligence, Inc.
1 Seneca St, Floor 24, Buffalo, NY 14203
privacy@getwi.com