Know when your resume gets seen.

Share your resume through Milo and see when it’s viewed, downloaded, and where the traffic came from. Never who.

Before the first interview

You are not bad at this. You are working blind.

Aarav, 2026 batch. Three projects, one hackathon, no internship yet. He has a resume he is proud of and forty applications he has heard nothing about. This is one month of his search.

Nobody warns you that the hardest part of your first job hunt is not the interview. It is the silence before one. You send the same document forty times and nothing comes back that tells you anything.

So you invent reasons. Not enough projects. The CGPA. No internship yet. The summary is weak. You rewrite it at midnight based on no evidence, because no evidence is all you have.

Maybe the resume was fine and nobody opened it. You would never know. That is the actual problem, and it is the only one Milo tries to solve.

A person working alone at a laptop late at night
Sending another application.
Sent
1
Replies
0
Times you were told why
0
  1. Applied. SDE Intern, 6 months.

What Milo solves today

Milo works on the link, not the file.

You share milo.app/l/abc123 instead of attaching the PDF. That link is a page Milo serves, so opening it, turning a page and downloading are all things Milo can actually see.

Attach the PDF to an email or upload it into a job portal and Milo sees nothing. A file that leaves your hands stops reporting back. There is a trick that would change that, embedding remote content inside the PDF so opening it pings a server, and Milo will not use it. Most readers block it anyway, and it is precisely the invisible tracking this product exists to avoid.

So this is the honest scope right now: Milo helps everywhere you control what you send.

  • LinkedIn messages and connection notes
  • Cold emails and follow-ups
  • Referrals, where someone forwards your link internally
  • Your portfolio site, GitHub README and email signature
  • Application forms with a "portfolio or resume URL" field

Portals that demand a file upload are outside what Milo can measure. Upload the file there, and use the link everywhere else. Follow what ships next.

Built in the open

See the full path →

The last few things that shipped, dated the day they landed.

  1. 31 October 2026

    2 releases
    • Shipped

      The account-wide activity log says which resume, and shows the whole event

      The /logs page lists events across every resume someone tracks, but each row just said "Page 1 viewed" with nothing identifying which one. ActivityLog now takes an optional linkTitleFor resolver, passed only by AccountLogs, since the per-link logs page is already scoped to one resume and doesn't need it. The row also surfaces every signal an event actually carries instead of a subset: utm_medium and utm_campaign were captured but never shown anywhere, and the relative timestamp ("2h ago") gets the exact time as a hover title. A raw session id was briefly shown here too and removed, not useful to a resume owner and an identifier with no reason to be on screen. Separately, an old event with no page_number, captured before this session's PDF-viewer rework added real page tracking, used to render literally "Page ? viewed"; it now reads "Resume viewed".

      • Frontend
      • Analytics
    • Shipped

      Forgot your password, without a way to probe who has an account

      POST /auth/forgot-password and POST /auth/reset-password reuse the registration flow's OTP mechanics, a hashed code, an attempt limit, a TTL-expiring record, but in their own password_resets collection, so a pending email verification and a pending password reset for the same user never collide on one row. forgot-password is written to be the one auth endpoint whose response never varies: no email echo, no timers, 200 whether or not the address has an account, whether it's verified, or whether an email actually went out, deliberately stricter than /auth/verify-email/status, which does leak known-vs-unknown and was an accepted gap for the less sensitive registration flow. reset-password follows the same rule, collapsing an unknown email, no pending reset, an expired code, and a wrong code into one password_reset_invalid rather than a response that would let the endpoint be used to test which emails have accounts. The frontend adds a two-step /forgot-password page and a show/hide toggle on every password field, shared by login and register.

      • Backend
      • Auth
      • Security
  2. 30 October 2026

    1 release
    • Shipped

      The resume viewer renders its own PDF, instead of trusting the browser's

      The /l/:slug page handed the fetched PDF straight to the browser's native viewer, an iframe pointed at the raw bytes. That viewer ships its own download/print controls, which save whatever's already loaded in the viewport without ever hitting /l/{slug}?download=true, so a click there was never counted as a tracked download event. Replaced by a canvas renderer built on pdfjs-dist: no native browser chrome, so Milo's own Download button (already wired to the tracked endpoint) is the only way to save the file. Page navigation and zoom are new controls this needed, and the dwell beacon now reports the page actually open instead of a hardcoded 1, the per-page tracking this feature always lacked for want of an in-browser renderer. A canvas has no DOM, so a link baked into the PDF (LinkedIn, GitHub, mailto:) isn't clickable on its own the way it was in a native viewer; restored by overlaying a transparent <a> per Link annotation, positioned from pdf.js's own annotation rects.

      • Frontend
      • PDF