Recagn
Sign in

Recagn does the screening. You make the calls.

Recagn turns a client's brief into a rubric, interviews candidates at times they choose and drafts the shortlist. Every decision waits for you.

Tuesday, 09:00Illustrative example

Senior Java Developer, Pune

  • RubricDrafted from the client's brief
  • Interviews21 of 30 done, at times candidates chose
  • Client reportDrafted, not published

Waiting for you · 2

This card changes as you try the page below.

Recagn runs screening from requirement to shortlist.

It does the work between the steps. The call at each step is yours.

  1. RequirementRead from the client's briefAnything the brief leaves open
  2. RubricCriteria, a gate and questions, draftedYou lock it
  3. Sourcing212 of 1,240 CVs read closelyYou choose who's invited
  4. InterviewsAt hours candidates choose, scoredYou read what's flagged
  5. ShortlistClaims checked, report draftedYou decide and publish

Recagn only had the brief.

You know the client.

Recagn writes the rubric. You decide what counts.

The brief said payment experience matters. On the call, the client told you what they meant. Choose the wording for criterion 2 and see what it changes.

What Recagn had · the client's brief

…You'll build Spring Boot services that move money, so experience with payment flows matters.

What you know · your note from the call, Monday

“They care that the person has handled reversals. Owning the service is a bonus.”

Rubric · 4 scored criteria, 1 gateDraft preview · rubric not locked
  1. 1Spring Boot services in production, 3+ years30%
  2. 2
    Choose the wording for criterion 2
    25%
  3. 3Kafka or RabbitMQ in production15%
  4. 4Code review and mentoring30%
  5. GNotice period ≤ 30 daysExample gate. The brief only said “immediate joiners preferred”, so a person sets the limit.gate
  6. Weights, added from the rows above100%

What this changes for Ananya R.

Draft preview · rubric not locked

Her answers don't establish service ownership. The draft criterion stays open.

3.35 of 5Recommended: Maybe

Same answers, same weights. Only the wording of criterion 2 differs.

  1. Criterion 2 · Recagn's draft

    Owned a payments or ledger service

  2. Her CV · line 4 · read before any interview

    “Payments experience: refunds and reversal flows for UPI”

    Line 4 says “payments experience”. It doesn't say she owned a service.unknown
  3. Her answers · all six questions

    Nothing in her answers says she owned the service. Not established, and not disproved.

    not established
  4. Criterion 2 score

    2/5

How 3.35 is worked out
Criterion and the answer it rests onWeightScore
Spring Boot services in production, 3+ years“I've run four Spring Boot services in production.” Q1 · 0:35304/5
Owned a payments or ledger serviceNothing in her answers says she owned the service. Not established, and not disproved. Q3 · whole answer252/5
Kafka or RabbitMQ in production“I mostly maintained the settlement consumers someone else built.” Q4 · 1:05153/5
Code review and mentoring“I pair with the two juniors on every pull request.” Q5 · 0:20304/5
Notice period ≤ 30 days“I can join in two weeks.” Q6 · 0:08gatemet

(4×30 + 2×25 + 3×15 + 4×30) ÷ 100 = 3.35Qualified from 3.5 · Maybe from 2.75 · on the interview scale, 1 to 5

Drafted by Recagn. Nothing counts until you lock it.

The wording you lock goes all the way to the client.

That's Recagn's draft wording. Change it ↑

Thirty minutes, on one of your roles.

Answer at a time that suits them.

Every candidate answers the same questions against the same rubric. They're told up front that AI helps evaluate their answers.

The clip sequence

  1. 0–4s. Question 3 of 6: Walk me through a payment you had to reverse in production. What did you check first? Under it: This interview uses AI to help evaluate your responses. You may request human review at any point.
  2. 4–13s. Ananya types her answer: We had a UPI double debit. I checked the reconciliation log first, then the idempotency key on the retry.
  3. 13–18s. She presses Save & next. Her answer is saved and question 4 of 6 opens: What have you done with Kafka or RabbitMQ in production?

Asking for a person takes one press.

A candidate can ask for human review at any point. Try the request as Ananya.

Illustrative interaction. Nothing is sent.

What the recruiter sees · 1 waiting

Dashboard → Needs attention

  • Lakshmi B.Human review requested

Try the candidate's request above. It arrives here.

Reading a request does not change the candidate's score or your Disposition.

Your client gets the shortlist, with the evidence.

Recagn drafts the report. You decide who is on it, and how long the link works.

Ananya R. Recagn recommends Maybe · 3.35 of 5

Shortlist · prepared for a payments company in Pune

Senior Java Developer, Pune

21 interviewed · rubric not locked yet

Selected1

  • Kavya N.Recagn recommended Qualified · 4.00 of 5 · decided by you at 09:12

Maybe1

  • Meera I.Recagn recommended Maybe · 3.15 of 5 · decided by you at 09:16

Not decided yet2

  • Ananya R.Recagn recommended Maybe · 3.35 of 5
    The evidence behind each score
    • Spring Boot services in production, 3+ years“I've run four Spring Boot services in production.” Q1 · 0:354/5
    • Owned a payments or ledger serviceNot established in her answers. Q3 · whole answer2/5
    • Kafka or RabbitMQ in production“I mostly maintained the settlement consumers someone else built.” Q4 · 1:053/5
    • Code review and mentoring“I pair with the two juniors on every pull request.” Q5 · 0:204/5

    CV check: Reversal handling: backed up. She walked through the double-debit fix step by step. Service ownership: not established. She fixed it on call and didn't say the service was hers.

  • Lakshmi B.Recagn recommended Maybe · 3.20 of 5

and 17 more who finished the interview

Recagn brings their history to the next role.

A new requirement starts with CV matches. When a candidate's earlier work applies, Recagn adjusts the order and shows why.

Next requirement

Java Lead, Pune

Spring Boot services · code review · up to 28 LPA

Rubric locked · shared Theme Key: backend

  1. 1
    Ananya R.Spring Boot · 4 years
    Match score 83History adjustment +5

    Her earlier interview backed up her Spring Boot experience. Shared backend criterion · halves after 135 days.

  2. 2
    Meera I.Spring Boot · 5 years
    Match score 84History adjustment 0

    CV match for this requirement. No history adjustment shown.

  3. 3
    Siddharth K.Spring Boot · 4 years
    Match score 82History adjustment 0

    CV match for this requirement. No history adjustment shown.

Her interview history is applied. Match scores stay the same.

Only the order changes. You still choose who to invite.

As the signal ages+5 → +2.5

Invented demo data. History belongs to this candidate and applies only where the locked rubrics share a Theme Key. Total adjustments are capped at ±15.

Recagn states what it does and doesn't.

How does Recagn score a candidate?

Against the rubric you locked. The weighted score is each criterion score times its weight, divided by the sum of the weights, on a scale of 1 to 5. See the working ↑

Can Recagn reject a candidate?

No. Recagn recommends. Only you can reject a candidate. See the decision ↑

What do candidates see?

They're told up front that AI helps evaluate their answers, and they can ask for a human review at any time. Try the candidate's screen ↑

What does a later role do with earlier evidence?

Recagn adjusts the sourcing order from that candidate's own earlier evidence: how they did on roles that share a criterion, your decisions and your client's outcomes, and whether they replied. The adjustment is capped. It shows the reason, and how fast it fades. It never changes a recommendation or an interview score, and it never invites or rejects anyone. See an adjustment

What happens to the details I give you for a demo?

We'll use this only to arrange your demo. We delete these details after 12 months without further contact. Privacy

See it run on one of your roles.