Ringo Work is an agent that builds a memory of your team's work. To do that it reads content from the tools you connect and keeps a durable, structured copy of it. This page describes exactly what that means, because "we take privacy seriously" tells you nothing.
1. Who we are
Ringo Work is operated by Launcher Capital. For any question about this policy or your data, contact contact@ringoai.app.
2. What we collect
Account information
Your email address and display name, received from Google when you sign in. We do not receive or store your Google password.
Workspace content
When you connect Slack, we read and store content from the channels you choose to include. The bot only sees channels it has been added to. This includes:
- Message text, including thread replies, in the channels you include
- Direct messages sent to the Ringo bot
- Reactions, and the meaning of your workspace's custom emoji
- File attachments. We do not store your files. Images, documents, and videos are converted to a text description, and we keep that description rather than the file itself.
- Channel and membership metadata, so the agent can respect who is allowed to see what
- Workspace member profiles, including display names and email addresses, which we use to match the same person across the tools you connect
Other connected tools
Everything else you connect — Jira, GitHub, Notion, and the rest — works differently, and the difference matters. We do not copy their contents into our database. The agent reads them on demand, within the permissions you granted, only when a task needs it. There is no bulk import and no background sync.
Two things from those tools can still persist. When the agent learns something durable — that a ticket is blocked, that its owner changed — it records that as a fact in your workspace's memory, and such a fact can name a person. And if an agent run fails partway, an excerpt of what it was writing can remain in our operational records. Both live until you delete the workspace.
Technical logs
For API requests we record the caller's IP address, user agent, and endpoint, to operate and secure the service.
3. Be clear about what we store
We store your message content, and we keep it. Ringo is a persistent memory product, not a transient one: content is saved in our database and used to build a knowledge graph of entities, facts, and relationships that persists between conversations. That is the product. If you need a tool that only reads messages in the moment and retains nothing, Ringo is not it.
When you first connect a source, we import at most the previous 30 days of history. From then on we ingest new content as it arrives.
4. How we use it
- To build and maintain your workspace's knowledge graph
- To answer questions and perform tasks you ask the agent to do
- To score which people, topics, and threads matter to your team
- To operate, secure, debug, and bill for the service
We do not sell your data. We do not use it for advertising. We do not use your content to train our own models, and we do not share it with anyone other than the processors below.
5. Subprocessors
These are the third parties that process data on our behalf:
| Processor | Purpose | What it receives |
|---|---|---|
| Railway | Hosting, database, agent compute | All data at rest and in transit |
| OpenRouter | Routes our AI model calls | Message content, embeddings, image data |
| Anthropic, OpenAI | Model inference, via OpenRouter or your own key | Message content and prompts |
| Stripe | Payments | Billing details only. No workspace content. |
| Pipedream | Tool connections, only for sources you connect through it | The agent's tool-call arguments, which may include content |
| Sign-in | Your email and display name |
6. AI processing and training
Answering your questions requires sending your content to AI models. Model calls are routed through OpenRouter, and we configure that routing to only use providers whose terms state they do not train on the data we send them. Customer content, including Slack data, is never used to train or improve any model, ours or a third party's. We also do not use Slack APIs to develop or train generalized AI or ML models.
We use models from OpenAI and Anthropic for extraction, embedding, scoring, and composing answers. Workspace admins can configure which model their agent uses, so the specific model processing your content depends on that setting. If you supply your own provider API key, your agent's calls go directly to that provider under your own agreement with them.
7. Retention
We retain your content for as long as your workspace exists. We do not currently expire content on a schedule; the memory is meant to be long-lived, and deleting it on a timer would defeat the product. You remove it by deleting it, as below.
8. Deleting your data
A workspace owner can delete an entire workspace from Settings → Danger Zone, or delete their whole account from account settings. Deleting a workspace removes its messages, its knowledge graph, its agent, and uninstalls the Slack bot. Deleting your account does this for every workspace you own.
Removing the Ringo bot from Slack stops further collection but does not by itself delete what was already collected — delete the workspace for that.
If you are a member of a Slack workspace but not a Ringo account holder, your workspace owner controls this data. Email contact@ringoai.app and we will work with you and your admin.
Depending on where you live you may have rights to access, correct, export, or erase your personal data, and to object to processing. Email us and we will act on it.
9. Security
Credentials — OAuth tokens, provider API keys, and client secrets — are encrypted at the application layer before they are written to our database. Workspace content and the knowledge graph are stored on our provider's encrypted volumes and are isolated per project, but are not separately encrypted at the application layer. All traffic is over TLS. Access to production is restricted to the engineers who operate it.
10. Browser storage
The Ringo web app keeps your sign-in session in a secure, HTTP-only cookie that browser scripts cannot read. It uses session storage for non-secret account display details and to preserve the page you intended to visit during sign-in. Local storage may retain interface preferences. Signing out removes the session cookie and session details; clearing the site's browser data removes the remaining local preferences.
11. Changes
We will update this page when our practices change, and will update the date above. For changes that materially affect how we handle your content, we will tell you directly rather than relying on you re-reading this page.