Follow-up emails
A follow-up email written for the stage you are actually at, not a generic template. Copy, edit, send.
Read onHiring almost never tells you what went wrong. The rejection loop does. Mark the job as denied, pick the stage it happened at, and the AI reads the JD plus your resume plus your keyword match and gives you three things: likely gaps, specific resume issues, and edits to make next time.
4 steps. Each one runs in seconds. Scroll in order.
One click from the tracker. A small form asks the stage (no response, after application, after phone screen, after interview, after offer) and lets you type a sentence about what you think happened.
JD, key skills, tailored resume, keyword match, any notes you left. It also reasons about the stage. A rejection after a phone screen means something different than a rejection after radio silence.
Likely gaps (skills the JD hunted for that your resume did not surface). Resume issues (phrasing or framing that may have dropped you). Recommendations (concrete edits for next time, not vague advice).
Tuned to the stage you reached. A nudge after silence reads nothing like a "thanks for letting me know" after a final round. The generator writes the draft; you edit it before sending.
One real example, same input reshaped two ways.
Senior Frontend role at a fintech. Stage reached: phone screen, then denied. Keyword match at time of application was on the lower end. Missing: GraphQL, Storybook, micro-frontends, A/B testing. User note: "I think it was the GraphQL question."
Likely gaps: 1. GraphQL appears in the JD and you flagged it in the call. The resume lists REST APIs only; adding a bullet about GraphQL usage, even basic, would have unblocked this. 2. Micro-frontends shows up in the JD too and is high-signal for a team explicitly rearchitecting. Resume issues: - Frontend bullets mention "APIs" generically. For fintech JDs, be specific. - No bullet ties work to compliance or regulated context. You did ship in a regulated environment (the master says so) but the tailored version dropped that. Recommendations: - Add one bullet naming GraphQL usage and the scale of it. - Restore the "regulated environment" line from your master. - For the next fintech posting, run the tailor again with these two edits baked into the master.
You should be able to tell how a feature works before you trust it with your job search.
This is the one step where model depth matters most. The heavier model reads the full record, reasons about stage context, and writes specific recommendations instead of boilerplate.
A rejection after submission is filter-driven. A rejection after phone screen is usually signal-in-the-conversation. A rejection after final round is usually close-calls or fit. The prompt carries the stage as a first-class field so the output matches.
If you have opted in to shared insights, the anonymized outcome data (job title, experience level, keywords present, stage) goes to the pool. No resume text. No company name. The opt-in is explicit on your profile page.
Each rejection analysis uses one of your three monthly AI operations on the Free plan, shared with resume tailoring and cover letters. Pro and Pro Weekly unlock unlimited AI ops on every job.
Honest limits read as trust signals. Hiding them does the opposite.
If your question isn't here, email support and you'll hear back from the founder.
No. It is a probability read. The AI reasons from the JD, your resume, the keyword match, and the stage. It names the most likely causes. Real hiring decisions involve factors we cannot see.
Mark it as "no response" if you got silence, or "after application" if you are not sure and only got the standard rejection email. The analysis adjusts its reasoning.
Only if you explicitly opt in to shared insights, and even then only anonymized metadata. Job title, experience level, keywords present, stage reached. No resume text, no company names, no personal identifiers.
You can, but the analysis is thinner. With a tailored resume the AI can reason about specific bullets. Without one it can only reason about the master vs the JD.
Sometimes. If the stage and context make a follow-up reasonable, the analysis suggests it and the draft generator writes the email. If not (final-round rejection after explicit feedback, for example) it does not push.
Every feature writes to the same record. These three are the closest neighbours.
A follow-up email written for the stage you are actually at, not a generic template. Copy, edit, send.
Read onAnonymized outcome patterns from the whole opt-in network, not just your own history.
Read onMatched and missing keywords named exactly. No opaque score, no black-box percent.
Read onCreate a free account in under a minute. First job tracked, first tailored resume, and first keyword breakdown all happen inside the onboarding flow.
No credit card · 2-minute setup · Cancel anytime