RIGEL
✦ For candidates

Get matched on what you actually know.

Upload your CV: Rigel turns it into a profile, scores it on skills, and shows you where you stand.

01

One CV, once.

Upload it, check what we read, confirm. That's the only thing the process asks of you.

02

Every role, automatically.

Confirm once, and every published role is scored against your profile — including every new one after.

03

No profile to perform.

No posts, no followers, nothing to keep current. Matching is a switch you can turn off at any time.

04

Feedback, not silence.

Which sections carry your score, which requirements keep coming up, and what isn't reflected in your profile yet.

01 · Profile

One CV. You keep the last word.

Most platforms ask you to re-type your CV into a form, and then ask again at the next company. Rigel reads the document once and turns it into structured evidence — the same shape for every role, so nothing has to be rewritten to fit.

The work isn’t applying. It’s rewriting.

Every role wants the same facts in a different shape. So the work of looking for a job isn’t looking — it’s rewriting one document, over and over, for software that decides in seconds whether a person ever sees it.

And the rewriting has a direction. You aren’t writing for the hiring manager. You’re writing for the filter standing in front of them — which is why so many people end up describing themselves in words they would never have chosen. Then the document goes in, and nothing comes back. The numbers below measure what that costs, and what the system asks of people to avoid it.

0.0%
of job seekers

believe searching for a job negatively affects their mental health and wellbeing.

0.0%
of job seekers

name waiting to hear back from employers as their single biggest source of job-search stress.

iHire Job Seeker Survey, 2024

The silence

0%

cite no updates or feedback after applying as a major frustration.

Monster, Application Black Box Report, 2026
0%

are frustrated by automated rejection emails that give no explanation.

Monster, Application Black Box Report, 2026

The labour

0%

are frustrated by re-entering résumé information they have already given.

Monster, Application Black Box Report, 2026

Writing for the filter

0%

modify their résumé to include job-description keywords most of the time.

Monster, Application Black Box Report, 2026
0%

have gone further — hidden text written to the filter, not to a person, to get a CV past it.

Greenhouse, 2025 AI in Hiring Report

The gap Rigel is built for

should employers be transparent about how they use AI?

0%

of job seekers say yes.

Greenhouse, 2025 AI in Hiring Report

do you trust AI to make faster and better hiring decisions?

0%

of hiring managers say yes.

Greenhouse, 2025 AI in Hiring Report
Step 1 of 2

Upload your CV

We'll convert your CV into a structured profile you can review and edit before matching.

Drag and drop your CV here
PDF or DOCX · Max 10 MB
or browse files
Upload and Process

Your CV is processed securely and only used for matching.

Upload it once.

Drop in a PDF or a Word file. Rigel extracts the text and reads it into the six sections every profile has: contact, summary, education, experience, skills and extras.


The structure is the point. A recruiter's criteria are structured too, so comparing one against the other is arithmetic rather than interpretation. That is why the profile has a fixed shape instead of a free-text box, and why you only have to build it once.

Contact

Name

Alex Carter

Email

alex.carter@hey.com

Title

ML Engineer

Phone

+1 (602) 555-0148

Location

Phoenix, AZ

LinkedIn

linkedin.com/in/alexcarter

Summary

ML engineer, six years in production computer vision. Led the model-serving rebuild that cut p99 latency 40%. Owns a model end to end — data, training, deployment and the

Skills
PythonExpertTensorFlowExpertComputer VisionIntermediateNeural NetworksIntermediateDeep LearningIntermediate
Experience
Senior ML Engineer
Northlight Systems · Mar 2021 – Present
· Rebuilt model serving on a batched inference path, cutting p99 latency 40%. Owned the vision pipeline from data collection through production monitoring.
Machine Learning Engineer
Cardinal Analytics · Jan 2019 – Feb 2021
· Built defect-detection models for high-volume manufacturing lines.
Education
Bachelor of Science in Electrical Engineering & Computer Science
University of California · 2011 – 2015
Machine Learning
Coursera (Andrew Ng) · 2021
Additional

No additional items extracted.

Check what it read.

Parsing is not magic, and Rigel doesn't pretend otherwise. Before anything is scored, the profile comes back to you as a draft — every field it extracted, in the words it extracted them.


What you read is what gets scored. The draft is the profile, not a preview of it — so anything wrong here is wrong in every match that follows.

Summary

ML engineer, six years in production computer vision. Led the model-serving rebuild that cut p99 latency 40%. Owns a model end to end — data, training, deployment and the

Skills
Python
Expert
TensorFlow
Expert
Computer Vision
Intermediate
Neural Networks
Intermediate
Deep Learning
Intermediate
Add skill
SaveCancel
Skills
PythonExpertTensorFlowExpertComputer VisionIntermediateNeural NetworksIntermediateDeep LearningIntermediate
Confirm ProfileRe-upload CV

Editing isn't a step.

Editing isn't a stage you pass through — it's available at both ends. Change anything on the draft before you confirm, and anything on your profile afterwards.


It matters most where the parser missed something the document already evidences. Alex's CV lists Machine Learning, Coursera (Andrew Ng) under education but names no skill from it, so PCA never reached the skills section. Confirming it is a correction, not an addition — the evidence was on the page the whole time.

Confirming is the gate: nothing is matched against any role until you do it. Upload a new CV later and the same loop runs again — a new draft, a new version, and the scores you already have stay attached to the version that produced them.

02 · Matching

You never apply. You’re always scored.

Confirming your profile is the last action the process asks of you. From then on, every role a recruiter publishes is scored against it — using the weights that recruiter set, by the same arithmetic applied to everyone else scored against that role.

Nothing to time, nothing to tune.

Everywhere else, applying is something you do, and doing it well is a skill of its own — the right words, the right file, the right day. That is a reasonable thing to get good at when the mechanism is hidden and nobody ever tells you what worked.


There is no Apply button.

Rigel has no such act. You confirm your profile once, and every role a recruiter publishes is scored against it automatically, by the same arithmetic used on everyone scored against that role.

The scoring is deterministic: the same profile, the same weights and the same scoring version produce the same number, every time. Nothing here is waiting to be timed, phrased or resubmitted. The only thing that changes a score is a change in the evidence itself.

Applications

Jobs where you’ve matched above the recruiter’s threshold.

8 applications3 strong matches1 shortlisted
Shortlisted(1)

A recruiter has shortlisted you. Full details unlocked.

Senior ML EngineerShortlisted
Vantibrook Systems
Phoenix, AZFull-timeApplied 10 May 2026
84Score

Vantibrook Systems is building large-scale foundation models for clinical decision support, with a focus on multimodal reasoning across imaging and structured EHR. We're looking for a senior engineer to lead model evaluation and own the serving pipeline end to end.Read more

Skills90%
Experience78%
Education84%
Decline this match
Other matches(7)

Recruiters can see you here, but haven't shortlisted you yet.

ML Platform EngineerSubmitted
ROLE #7586Applied 12 May 2026

Company and full job details unlock when you’re shortlisted.

71Score

Where you stand.

When your score crosses the threshold a recruiter set for their role, an application appears in your list. You didn’t send it — the score created it, and the recruiter sees you ranked among everyone else who crossed the same line.


What you can see depends on where the match has got to. Until a recruiter shortlists you, you see the role and your score, and nothing that would tell you whose role it is. That asymmetry is deliberate: it is the same wall that keeps your name, your photo and your contact details off their screen while they are ranking you.

  • The role title and your overall score. From the moment the score is created.
  • Nothing about the company. Not the name, not the description, not the location — until they shortlist you.
  • The sections behind the score, once you’re shortlisted. Skills, experience and education, each with its own number.
  • A way out at any point. Decline a match and it stops being visible to them.

None of this depends on you checking. A role published on a Tuesday is scored against your profile on Tuesday, whether you’re looking or not — and if it crosses the line, the first you hear of it is that it’s already yours.

03 · Control

Nothing to keep up.

Once your profile is confirmed, none of this asks anything of you — and none of it happens without your say. You can stop it at any point, and stopping is a switch, not a request.

You don't need 500 connections. Just yourself.

A headshot. A follower count. Posts to keep up, a network to grow, a profile that has to look busy. Everywhere else that work is the price of being seen — not because any of it says whether you can do the job, but because visibility has to be rationed somehow and that’s the mechanism on offer.


None of it is an input here. Your score comes from your evidence measured against the weights a recruiter set, and there is nowhere in that arithmetic for a follower count to enter.

You’ll still open Rigel — for a new match, for a recruiter’s shortlist, for what your scores are telling you. But you come back to read something, not to stay visible. The only thing you can do that changes a score is change the evidence, which is the point.

Applications

Jobs where you’ve matched above the recruiter’s threshold.

8 applications3 strong matches1 shortlisted
Shortlisted(1)

A recruiter has shortlisted you. Full details unlocked.

Senior ML EngineerShortlisted
Vantibrook Systems
Phoenix, AZFull-timeApplied 10 May 2026
84Score

Vantibrook Systems is building large-scale foundation models for clinical decision support, with a focus on multimodal reasoning across imaging and structured EHR. We're looking for a senior engineer to lead model evaluation and own the serving pipeline end to end.Read more

Skills90%
Experience78%
Education84%
Decline this match
Other matches(7)

Recruiters can see you here, but haven't shortlisted you yet.

ML Platform EngineerSubmitted
ROLE #7586Applied 12 May 2026

Company and full job details unlock when you’re shortlisted.

71Score
Machine Learning EngineerSubmitted
ROLE #4430Applied 9 May 2026

Company and full job details unlock when you’re shortlisted.

66Score
ML Engineer, PerceptionSubmitted
ROLE #5991Applied 8 May 2026

Company and full job details unlock when you’re shortlisted.

63Score
app.rigelsignal.com says

Decline this match? It will stop being visible to this recruiter. This can't be undone.

CancelOK

You can turn it off, or turn any match down.

Matching is a switch in your settings. Turn it off and you stop being scored against new roles; turn it back on and it resumes. Nobody is notified, and you don’t have to say why.


The same applies one match at a time. Decline a match and it stops being visible to that recruiter — no message, no explanation, no reason to give.

And there is no document to keep current. One CV, once. Nothing since has asked you to update it, tailor it, or send it anywhere.

04 · Insights

Every match teaches you something.

Every score is kept — the few that crossed a threshold and the many that didn’t. Because the scoring is arithmetic rather than opinion, all of it can be counted: which sections carry you, which requirements keep coming up, how the picture changes as your profile does. No recruiter was ever going to write you that.

The scores below the line still count.

Most roles you’re scored against won’t cross their threshold, and you’ll never see them as applications. They aren’t thrown away. Every score is stored with the weights and the scoring version that produced it, so a match from three months ago still reconciles today.


Together they make a picture rather than a verdict: which sections carry your score, which skills recur in the roles where you score well, and which ones aren’t currently reflected in your profile. It describes where you are. It is not a target to hit, and Rigel will never tell you to add a skill to raise a number.

None of it is a verdict on you. It is a record of how your evidence met a series of specific asks, kept so you can read it rather than guess at it.

See it in action

Watch a CV become a match.

From the CV you upload to a confirmed profile to the score behind every role: the whole flow, start to finish.

rigel · cv → match walkthrough
00:00 · Upload
Your CV becomes a profile
00:52 · Match
Scored against every role
01:40 · Explain
The math behind your score

See where you stand.

Upload your CV. Get matched on every relevant role automatically.

Apply with clarity →