Users can initiate resume tailoring from any job posting in the job board by clicking Tailor Resume.
The resume engine uses the Anthropic Claude API to read profile entries and rewrite a resume of set format to fit the posting’s job description.
The tailored resume data is validated before being rendered to PDF.
Resume tailoring will use placeholder content if no user-created profile entries exist.
Resume tailoring is locked if the Claude API key is absent or unreachable.
Every AI call made during tailoring is tracked with token counts and estimated cost.
The user can request a fresh tailor pass on a posting they have already tailored, generating a new resume variant (Re-tailor).
Templates
The validated data is rendered into a source file using a user-selected named template.
At least two templates are shipped: classic and modern.
Compilation
The source file is compiled to PDF using the configured PDF compiler.
Resume compilation is locked if the PDF compiler is not found.
On compilation failure, the most actionable error is surfaced to the user.
Preview and Snapshot
The compiled PDF is displayed natively inside the app.
A snapshot of the resume data is stored alongside the source file.
If the source file is lost, the PDF can be recompiled from the stored snapshot (Recompile from snapshot).
Resume Naming
Each tailored resume record has a user-editable display name (name), separate from the posting title.
The name is edited inline in the resume list sidebar: clicking the current name switches it to a text input; pressing Enter or blurring commits the change; pressing Escape cancels.
The name defaults to empty; the sidebar falls back to displaying the posting title when no name is set.
Tailoring a resume does not record an applied date; applied_at is nullable and is set only when the user explicitly marks the posting as applied.