Privacy Policy
This explains what FunctioningHQ holds, why, who else sees it, how long it is kept and what you can ask of us. For the support conversations a workspace connects, the business that owns the workspace decides what happens to them, and we process them on its behalf.
1. Who we are, and our two roles
FunctioningHQ is made and run by Functioning Labs LLC, Boston, MA 02135 (“we”, “us”).
As a processor. When a business connects its helpdesk, the AI it built, its widget, inbox or files to a workspace, that business is the controller of the personal data in them, and we process it only on its instructions: the settings, connections and approvals it chooses, and the data processing agreement it accepts in the product. If you are one of that business’s customers, ask the business about your data first; we will help it answer.
As a controller. We decide what happens to the account details of the people who sign in to FunctioningHQ, to what this website records, and to what you give us through the demo or the contact page.
Questions about privacy go to privacy@functioninghq.com.
2. What a workspace holds
What comes in depends on what the business connects. It can include:
- Support conversations and their details (when, which channel, the topic, tags, ratings, whether the AI or a person answered, and what the helpdesk reported about resolution), redacted on arrival.
- A pseudonym for each customer: a salted hash of the identifier the source gave, so people can be counted and followed across conversations without their identity being held. Company names are kept, so issues can be grouped by account.
- One send address per person, only if the business turns it on, for sources that carry no contact record (a widget, an imported file or its own key). It is encrypted, read only at the moment an approved message is sent, and shown on no screen.
- The address somebody gives for a reply, when they write to the business in its widget and ask to be answered by email. It is kept encrypted with that one conversation, whatever the send-address setting says, read only when the team’s reply is sent, deleted with the conversation, and included in an export of that person.
- What the service works out: verdicts on answers with the passages they cite, patterns, drafted fixes, measurements, reports and the nightly brief.
- Answers to surveys and feedback people choose to give, redacted like everything else. Open survey links are anonymous by design.
- The workspace’s own documentation: its help center, internal articles, and public pages or changelogs it asks us to read.
- About the team: the helpdesk’s identifier for the teammate a conversation was assigned to, hashed like a customer’s, used only for one person’s coaching and never to rank anyone; and for people invited to training, a work email kept encrypted to send them their link.
Raw helpdesk payloads are never stored. The detail of what each connection reads is on Security and in the product, before you connect.
3. What redaction removes
Before anything is written, the customer’s name, email and phone number as the source holds them are removed wherever they appear, and so is anything shaped like an email address, phone number, card or bank account number, national identifier, postal address, API key, token or credential; each is replaced with a placeholder. Text that arrives later, such as a survey comment or a note, goes through the same redaction. Redaction is automatic and very good, not perfect: a business can add its own patterns, and we fix any gap we are told about.
4. What we hold as controller
- Account details of the people who sign in: name, email address, the workspaces you belong to, your role and permissions, and your notification choices. Sign-in itself is run by our authentication provider.
- What you do in a workspace, in its audit trail: who approved, changed or exported what, and when. Our own staff’s access is recorded the same way, with the reason they gave.
- Billing status: the plan, whether the subscription is active, and the amounts. Card and billing details stay with Stripe (see Paying for it).
- Messages you send us through the contact page or by email.
- Error reports when something breaks, with request bodies, query strings, cookies, the signed-in person and customer text removed before they leave the app.
- Before you are a customer, if we test a prospect’s public AI on our own initiative: the conversations we had with it and its public pages, deleted after 30 days.
5. Why we use it
Workspace data is used only to provide the service the business configured: to read and check its conversations, find what keeps going wrong, draft and carry out the fixes it approves, measure them, and keep the service secure. Nothing else.
What we hold as controller is used to run your account and bill for it (to perform our contract with you), to keep the service secure and fix faults, to answer you, and to improve the product from counts about how it is used (our legitimate interests), and to send you product updates only if you asked for them (your consent, which you can withdraw at any time). We do not sell personal data or make decisions about anyone that have legal or similarly significant effects on them. Our public website tells Google and LinkedIn which of its pages you opened, to measure our ads, only as the cookies section below describes; nothing from a workspace, the demo, sign-in or sign-up is ever shared for advertising.
6. AI models
FunctioningHQ uses AI models to read conversations, check answers, group problems, draft fixes and answer questions in the assistant. Every model request goes through one gateway to Anthropic or OpenAI, carries redacted text only, and asks for zero data retention: the provider keeps no prompt or answer once it is returned (OpenAI keeps a narrow exception for safety review), and the gateway keeps none either. Neither trains on it, and we never use your data to train models.
Search by meaning is off unless a workspace turns it on. When it is on, a short summary of each conversation (never the conversation itself) is turned into numbers that help find similar conversations; turning it off deletes them the next night.
7. Who else sees it
Our subprocessors, each for one job and bound by a contract to protect the data: they are listed, with what each receives and where, on Security. We tell workspaces before one is added.
The systems a business connects (its helpdesk, Slack, Linear, Notion and the like) are its own, and receive only what it approved for them. We may also disclose information if the law requires it, to protect people’s safety or our rights, or to a buyer of our business, which would be bound by this policy. Our staff open a workspace’s data only with a reason they record first, for a limited time, and the business sees every opening on Settings › Activity.
8. Paying for it
9. Where it is processed
Each workspace’s database is in the United States or in the European Union (Frankfurt), chosen when the workspace is created; it cannot be moved afterwards. Model requests and some of our providers are in the United States. Where personal data leaves the EEA or the UK, we rely on the Standard Contractual Clauses.
10. How long it is kept
- Conversations: for the retention window the business sets: by default as long as its plan keeps history (1 year on the paid plan, 90 days on Free), shorter if it chooses, and never under 7 days. Each night, conversations past it are deleted with everything derived from them, and deleted data leaves the database’s restore history within 6 hours.
- One person: a business can erase a person across every source at any time.
- A closed workspace: destroyed with everything in it 7 days after it is closed, and a receipt sent.
- A free workspace that does not choose a plan: deleted after its free period, with an email first; its report stays readable for 90 days.
- The demo: the copy made for you is deleted within two hours.
- The audit trail and access log: for the life of the workspace, because they are the record of who did what. Someone who leaves is shown as a former member.
- Account details: while you have an account, and then deleted, except what we must keep for tax, accounting or legal reasons.
- Error reports: for our error-tracking provider’s standard period, then deleted.
11. The demo
13. Security
14. Your rights
Depending on where you live, you can ask to access, correct, delete or receive a copy of your personal data, to object to or restrict how it is used, and to withdraw consent you gave. Residents of California and other US states with privacy laws have these rights too. We do not sell personal data; the advertising tags on our public website may count as sharing it for advertising under some of those laws, and Your privacy choices at the foot of every page turns them off. We honor Global Privacy Control as the same request. We will not treat you differently for using these rights.
If you are a customer of a business that uses FunctioningHQ, it can export or erase your data itself, and it is the right place to start; if you write to us instead, we pass your request to it. For anything else, write to privacy@functioninghq.com. We answer within 30 days, usually much sooner, and may need to confirm who you are first. In the EEA or the UK you can also complain to your data protection authority.
15. Children
FunctioningHQ is a service for businesses and is not meant for anyone under 18. We do not knowingly collect personal data from children; if you believe a child has given us some, write to us and we will delete it.
16. Changes to this policy
When this policy changes, we update the date above. If a change matters, such as a new use of data or a new kind of recipient, we tell workspace admins by email before it takes effect. A new subprocessor is announced before it is used, and each workspace is asked to accept the new list.