Most résumé tools ask you to trust a black box: your work history and the job posting you're chasing go in, and somewhere behind the scenes a model neither you nor the tool's marketing page will name reads it, processes it, and maybe keeps it. You never see which model, who pays for the call, or where the data sits afterward.
We think that's backwards. A tool that's rewriting your career history should let you pick the brain doing the rewriting — and should never need to hold onto your key to do it. That's the whole idea behind bring-your-own-key (BYOK): you connect the AI account you already trust, and Tailr stores it encrypted and uses it only to run your tailoring requests.
What "bring your own key" actually means
When you connect a provider, your key is saved to your Tailr account and encrypted at rest with AES-256-GCM — never in plaintext, never in a log file. Because BYOK is an account-based plan, it's stored once rather than re-entered each session and follows you across devices; Tailr uses it only to run your tailoring requests against the model you chose, and nothing else. It's wiped the moment you remove the connection or delete your account.
You pick the model — and the cost
Tailr connects to OpenAI, Anthropic's Claude, and Google's Gemini, plus roughly ten more OpenAI-compatible providers, so you're not locked into whichever model happens to be cheapest for us to run. Maybe you already have OpenAI credits from another project. Maybe you prefer Claude's writing style, or you're testing a smaller, faster model for quick drafts. That's your call, and it's also your bill — you pay your provider directly, at their rates, with no markup sitting in between.
This matters more than it sounds. A tool that routes every user through one shared backend model has an incentive to pick whatever's cheapest for its own margins, not whatever writes your bullets best. BYOK removes that conflict entirely: the model recommending your next sentence answers to you, not to someone else's cost sheet.
You choose the model and bring the key — we keep it encrypted and use it only to run your work.
No key? You still get real tailoring
Not everyone wants to manage API keys, and that's fine. If you don't connect a provider, Tailr falls back to a managed server key on the free tier, so you still get genuine résumé tailoring — not a stripped-down demo. Bringing your own key isn't a paywall you have to clear to use the product; it's an option for people who want more control over model choice, cost, or where their data flows.
Connecting a key takes under a minute
- Open Settings and choose a provider — OpenAI, Claude, Gemini, or any of the compatible options.
- Paste the API key from that provider's dashboard.
- Pick a model from the list Tailr fetches for that provider.
- On connect, Tailr saves your key to your account, encrypted at rest — automatically, no extra step.
- Tailor as usual — every generation now runs on your key and your chosen model.
The rest of your data gets the same treatment
Key privacy is one piece of a bigger habit. When you upload a résumé, it's parsed in memory and never written to disk. When you paste a job posting instead of uploading a file, nothing about that changes. And if you decide Tailr isn't for you, self-serve account deletion wipes your data — including any encrypted key you'd chosen to save — rather than leaving it to expire on its own schedule. None of this is a feature you have to dig for in a privacy policy; it's just how the pipeline is built.
Why this is worth the extra setup step
Connecting a key is a small amount of friction compared to typing an email and password. We think it's worth it because it hands you two things most AI tools quietly keep for themselves: which model reads your career history, and what that reading costs. Gaps in your résumé are information, not shame — the same principle applies to the tool itself. You should be able to see exactly what's touching your data and change it whenever you want, not take it on faith.