Privacy policy
Draft, pending legal review. Last updated 25 October 2026.
Tkay helps you turn something you learned into a small, checked try on your own project. To do that we keep what you capture, the projects and goals you set, and what the crew makes for you. We don’t sell your data, we don’t show ads, and we don’t track you across other apps or websites.
This policy covers the Tkay mobile apps and the services behind them, run by UpLevel Technologies LLC (“we”). Questions go to chris@uplevel.llc.
What we collect
| Data | What it is | Why |
|---|---|---|
| Email address | The address you sign in with. With Sign in with Apple this can be Apple’s private relay address. | Your account, sign-in codes and replies to support requests. |
| Name | Only if you share it through Sign in with Apple. It’s optional. | Greeting you and showing who did what in a shared workspace. |
| User ID | An identifier we create for your account. | Keeping your workspaces and records together. |
| Photos and videos | Pictures you choose to capture or attach. We read text in them on our own servers so the crew can use it. | Turning what you captured into ideas and drafts. |
| Audio | Voice notes you record. We transcribe them on our own servers. | Same as above. |
| Other content you add | Notes, links and files you capture, plus the projects, goals, plans, results and drafts in your workspace. | Running the app: matching ideas to goals, planning tries and showing results. |
| People you invite | The email address of someone you invite to a shared workspace. | Sending the invitation and adding them when they accept. |
| Connected services | Only if you connect GitHub or Linear: your account name and ID there, which repositories or Linear workspace you chose, and, for Linear, access tokens, which we store encrypted. We don’t keep a GitHub password or personal token. | Reading what you asked Tkay to read, and opening a draft pull request when you approve one. |
| Support messages | What you send through Help and support, including reports about content. | Answering you and handling safety reports. |
| Purchase history | What you bought in the app, as reported by Apple or Google. We never see your card details. | Giving you what you paid for and restoring purchases. |
| Product interaction | Actions in your workspace, such as approving a plan or keeping a draft, with timestamps. | Showing your history, enforcing approvals and keeping an audit trail. |
When you sign in we also keep a session record with the IP address and device type it came from, so we can keep your account secure. We don’t use a crash-reporting or analytics SDK, and we don’t collect your location or contacts.
All of the above is linked to your account. None of it is used for tracking or advertising.
On your phone
Captures are saved on your device first, in an encrypted database, so nothing is lost when you’re offline. The encryption key is kept in your device’s secure storage (Keychain on iPhone, Keystore on Android). Deleting the app removes this local copy.
Links you capture
When you capture a link, our servers fetch that page directly, so the website sees a request from Tkay, not from your phone. We keep a copy of the page with your capture. Takeaways are pulled out of it on our servers with Tkay’s own rules; the page isn’t sent to an AI model.
Links to GitHub repositories are read through GitHub’s API instead: the repository’s description, topics, language, star count, licence, top-level file names and its README (or the release notes or file you linked). A public repository is read without any account. A repository that someone in your workspace connected to Tkay is read with read-only access to that one repository. For a public repository, what was read is sent to OpenAI to write takeaways: the repository name, description, topics, language, star count, licence, top-level file names and README. Each takeaway has to quote the README exactly or it’s dropped. A private repository’s README is never sent to OpenAI from a link.
AI and what is sent to OpenAI
Tkay uses OpenAI’s models for the jobs below. Each one starts only when you do the thing that starts it, and only what that job needs is sent. Anything AI wrote is labeled as AI-written in the app.
- Drafts. Before you approve a plan, it tells you whether AI will write the draft. When it does, we send the idea, your goal, your starting point, the relevant text from what you captured, and what your project already records about it: earlier tries and their results, your rules and missing prerequisites. If you try the idea on a code repository, the relevant files and the repository’s list of file names are sent too.
- Takeaways from GitHub links. When you capture a link to a public GitHub repository: its README and public details, as described above.
- Suggestions from a repository. When you ask Tkay to fill in a project from a repository, or to set up projects from your repositories: the files and details listed under GitHub below.
- Suggestions from Linear. When you pick a Linear project to draft from: its name, description, status, target date, teams and milestones, and up to 50 open issues with their titles, descriptions, priority, state and labels.
- Conference agendas. When you import an agenda: the page’s text, unless the page already lists its talks in a structured form Tkay can read without AI.
Turning your notes, files, photos, voice notes and other links into takeaways happens on our servers without AI. Their text reaches OpenAI only as part of a draft you approve. Matching ideas to your goals doesn’t use AI either.
OpenAI processes this for us under its API terms and does not use API data to train its models by default.
When you first set up Tkay you agree to this processing, and we ask you again whenever what we send, or to whom, changes. Our servers check that agreement each time one of these jobs is about to send something to OpenAI, so nothing is sent without it, including from an older version of the app. Without it, captures are still read on our servers without AI (a GitHub link’s takeaways are quoted straight from its README), and the jobs above ask you to agree before they run.
GitHub, only if you connect it
If you connect GitHub, you choose which repositories Tkay can use. Tkay copies a repository into an isolated sandbox for a try, runs its checks, and deletes the copy afterwards. These sandboxes run at Daytona, which receives the copy of the repository for that try only; they hold no Tkay or GitHub credentials and are deleted when the checks finish. It opens a draft pull request only after you approve that step. We keep the repository names you chose and the results of each try, including the changes it proposed, but not a full copy of your repository. You can disconnect your GitHub account in the app, and remove Tkay’s access to repositories in your GitHub settings, at any time.
If you ask Tkay to fill in a project from a repository, or to set up projects from your repositories, it reads a small part of each repository at its latest commit: the README, contributing guide, package files, CI settings, documentation headings, recent commit messages and open issue titles. It never reads files that look like secrets or keys, and it runs and changes nothing. That text is sent to OpenAI to draft suggestions, which you review before anything is added. This applies to private repositories you connected too. We keep the suggestions, the list of files read and the commit they came from, not the text itself.
Linear, only if you connect it
If you connect Linear, Tkay gets read-only access and never changes anything in Linear. It reads your Linear name, email and user ID and your Linear workspace’s name (we keep the name, the ID and the workspace name), your teams and most recently updated projects, and, for a project you pick, its details and up to 50 open issues. We keep a dated copy of what it read, with the suggestions drafted from it, in your workspace. When you ask for suggestions, that project’s text is sent to OpenAI as described above. Disconnecting revokes Tkay’s access at Linear and deletes its tokens, and so does deleting your account.
Conference agendas
When you import a conference agenda, Bright Data fetches that one page for us from its own network, so agenda pages that block automated visits still load. Bright Data receives the page address you gave, not your account details. If the page doesn’t list its talks in a structured form, its text is sent to OpenAI to find them. Talks you keep include the speakers’ names and affiliations as the organiser published them.
Who processes data for us
These companies process data on our behalf, only to run Tkay. GitHub, Linear, Apple and Google are also services you choose to use with Tkay, so their own privacy policies apply to what you keep with them.
| Provider | What they do for us | What they receive, and when |
|---|---|---|
| Railway | Hosts our servers, our database and the storage for files you upload, pages you capture and exports. | Everything Tkay keeps, all the time. |
| Cloudflare | Serves this website and runs our domain names. The website has no analytics and sets no cookies. | Standard request details, such as your IP address, when you visit tkay.app. |
| OpenAI | Writes AI drafts, takeaways from public GitHub READMEs and project suggestions from a repository or Linear project, and reads conference agendas. | The text each job needs, listed above, only when you start that job. |
| Bright Data | Fetches conference agenda pages. | The address of the agenda page, when you import one. |
| Resend | Sends sign-in codes and workspace invitations by email. | The recipient’s email address and the message, when you ask for a code or invite someone. |
| Apple | Sign in with Apple, and purchases on iPhone and iPad. | What’s needed to sign you in, confirm purchases and, when you delete your account, revoke Tkay’s sign-in. |
| Purchases on Android through Google Play. | What’s needed to confirm and restore purchases. | |
| GitHub | Reads repositories you link or connect, and opens draft pull requests you approve. | Requests for the repositories involved, when you capture a GitHub link, connect GitHub or run a try. |
| Linear | Reads your Linear projects, only if you connect it. | Read requests, when you browse or draft from your Linear projects. |
| Daytona | Runs a repository’s checks in a temporary sandbox when you try an idea on it, then deletes the sandbox. | A copy of the repository at the commit being tried, with the proposed change, for that try only. |
We don’t sell or rent personal data, and we don’t share it with advertisers or data brokers. We disclose data to others only if the law requires it, to protect people from harm, or as part of a merger or sale of the business, in which case this policy continues to apply.
Shared workspaces
Your personal workspace is private to you. If you create a shared workspace and invite people, they can see what’s in it according to the role you give them. If you’re invited to someone else’s workspace, its owner controls that content.
How long we keep it
- Your content and account data stay until you delete them or delete your account.
- When you delete your account, we delete your workspaces, captures, files, drafts, connected memory and sign-in records. This finishes within 30 days.
- Deleting your account also revokes Tkay’s access at Linear and its Sign in with Apple authorization. You can remove the Tkay GitHub App from your repositories in your GitHub settings.
- We keep a minimal record that the deletion happened, plus security audit entries, without your content. We keep purchase records as long as tax and accounting laws require.
Your choices and rights
- Get a copy: Account → Your data → Export my data.
- Delete your account: Account → Your data → Delete my account. You can also ask us by email.
- Withdraw your agreement to AI processing: Account → Your data → Withdraw AI consent. From then on nothing is sent to OpenAI for you, including by work you started earlier that hasn’t reached that step yet, and the rest of Tkay keeps working without AI. You can agree again in the same place. Withdrawing doesn’t undo processing that already happened. You can also email chris@uplevel.llc.
- Correct or ask a question: email chris@uplevel.llc.
Depending on where you live, you may have more rights, such as objecting to processing or complaining to a data protection authority. We honour these requests wherever they apply. We reply within 30 days.
Security
Data is encrypted in transit. On our side, each workspace is isolated at the database level, and server credentials are kept in a password manager, never in the app. No system is perfectly secure; if something goes wrong that affects you, we’ll tell you.
Children
Tkay is not meant for children under 13, and we don’t knowingly collect their data. If you believe a child has an account, email us and we’ll delete it.
International transfers
Our providers may process data in the United States and other countries. Where required, we rely on standard contractual clauses or equivalent safeguards.
Changes
If we change this policy in a way that matters, we’ll tell you in the app before it takes effect. If the change affects what we send to OpenAI or other companies, the app shows what changed and asks you to agree again before you keep using it. The date at the top shows the latest version.