Stop praying the client's interviews catch your weak ones.
Right now you send fifty and hope. Score them first and send eight you'd stake your name on and stop outsourcing your filtering to the client who's judging you for it.
For recruitment agencies
Recruitment Software for Agencies That Helps You Send Better Candidates, Not More Resumes
Score candidates against the real must-haves before you submit, and build shortlists you can confidently stand behind with a receipt the client can read line by line. Paraakh with CheckedIn gives growing agencies the capabilities expected from the best ATS for small recruitment agencies and the best recruiting tools for small businesses.
8scored · submittedwas 50 unscreened
Priya S.91submitted
Kavya R.87submitted
Imran Q.62held back
Before anything else
The submission you lose to a duplicate costs you 100%, not a margin.
You sourced, screened, and submitted. Someone else submitted the same person three days earlier. You get nothing, after the cost was already spent. Every other inefficiency on the desk is a percentage. This one is the whole fee.
First submission You own the candidate
For the window in your MOU. Ninety days, or whatever you agreed. Nothing automatic can reassign it inside that window: no re-import, no bulk action, no admin shortcut. And the server stamps the submission time, not the submitter, which cuts both ways and is exactly why the record holds up.
Second submission You are told, immediately
Owned by another partner until 12 Oct. Not a silent rejection, not a mystery, not a race to complain loudest. You find out before you spend the hours, so you can move the recruiter onto a mandate you can actually close.
On joining The invoice reads the ownership record
Not whoever filed the paperwork. At the version of your fee terms in force at the time, so a renegotiation next quarter cannot rewrite this placement.
On a dispute You can produce the submission order
Who submitted, when by the server clock, who won, until when. An append-only log rather than a field somebody updated, which is what makes it usable in an argument.
At renewal Your scorecard with that client
Submissions, pass rate, submission-to-placement, TAT. Evidence to put on the table when they push 8.33% downward. Yours only, never another agency's.
CheckedIn + Paraakh
Drop CVs in a folder. Get a client-ready shortlist back.
CheckedIn runs the grind end to end: it reads every CV, parses it into a structured profile, scores it against your client's JD, and hands you a ranked shortlist with a receipt. Paraakh measures the skill where you need proof rather than a claim. You place. It chases.
01 Ingest
Drop resumes in the role's folder or forward them any format, in bulk.
02 Enrich
It parses each resume into a structured profile skills, seniority, stack, and the gaps.
03 Assess
Auto-builds a profile-aware test + AI voice screen, proctored. Real skill, not claims.
04 Match
Scores every candidate against your client's JD fit, gaps, and the why.
05 Deliver
A ranked, client-ready shortlist with evidence lands back in your inbox.
MANUAL TODAY
Read 200 CVs, test them one by one, and hope the client's interviews catch the weak ones.
WITH CHECKEDIN
Drop the folder. Send eight, each scored against the JD, each with a receipt the client can read.
Honest about the doors: CheckedIn does not post to Naukri or LinkedIn and does not search Resdex. It screens and ranks the people you already have. Keep sourcing where you source, then export the sheet or forward the mail and it lands in one funnel with a per-row report of created, duplicate and errored.
The number that decides everything
Submit fifty to land one, and you're the agency that sends junk.
Land one in five, and you're the partner the client trusts and pays a premium for. The only thing standing between those two reputations is knowing before you hit submit who's actually a fit.
The arithmetic nobody puts on a slide
A recruiter's month is a fixed cost. Their output is submissions. So your submission-to-interview ratio is not a soft metric, it is the price of every interview you generate, and you can move it without hiring anyone.
Loaded recruiter cost₹45,000 / month
Submissions in the month30
Cost per submittal₹1,500
At 1 interview per 10 submissions₹15,000
At 1 interview per 4 submissions₹6,000
Illustrative. Substitute your own loaded cost and ratio.
Built for how hiring actually works here
Notice period is scored
One of only seven factors. A 90-day notice against a role that needs somebody in 30 shows as a gap on the receipt rather than a surprise in week six.
₹18,00,000 reads as ₹18 lakh
Not eighteen thousand. Indian digit grouping and CTC formats are parsed properly, because a salary factor that misreads the number poisons the whole score.
Bulk from the boards you already pay for
Export from Naukri, drop the sheet in, get a per-row report. No board integration is claimed, because there is none.
Empanelment, not just "onboarding"
A client approves your connection, then grants access to specific roles. Revoking it is recorded, and you keep the candidates you sourced.
How your desk changes
Fewer profiles out. More placements in.
CHECKEDIN + PARAAKH
Your business and your proof, working together.
CheckedIn runs the desk: candidates, mandates, pipeline, clients, ownership and invoices. Paraakh measures whether a candidate can actually do the job. The join between them is the part that matters, because a measured score doesn't sit in an attachment beside the CV. It replaces the claim inside the ranking.
◆
CheckedIn owns the record
Who submitted first, who moves, which client, which stage, who gets paid.
◆
Paraakh owns the measurement
How good a candidate actually is, against one specific role.
Arjun M.Skills-verified · Java / Spring
Core Java88
Spring Boot81
Databases76
MEASURED, NOT CLAIMEDREPLACES THE CV
Illustrative. Bars show measured per-skill results, not CV keywords.
What moves
The numbers an agency lives or dies on.
Fewer profiles, more interviewsSend the ones who get the call
More placements per mandateThe number your business lives on
Less time screeningScored on arrival, not one by one
Clients who keep callingDefensible quality keeps the mandate
Paraakh assessment workspace · multi-client, white-label
Serve every client from one workspace. Each one isolated, each one branded.
Run a hundred mandates side by side without spreadsheets, shared accounts, or "which client was this candidate for again?" Every client gets a workspace that looks like yours your logo, your colours, your domain with their candidates, their results, and nobody else's.
To be precise about which product does what: the white-labelling and per-client workspaces below belong to the Paraakh assessment platform. CheckedIn's model is different by design: one workspace for your agency, connected to each client who empanels you, with access granted per role rather than to everything.
🏢 A workspace per client
Hard isolation at the data layer your client A never sees client B's candidates, even by accident. Even your recruiters can be scoped to specific tenants.
🎨 Your brand, not Paraakh's
Custom logo, primary colour, favicon, email templates. The candidate sees your agency. The client sees your agency. Paraakh stays invisible.
🔐 Permissions per recruiter
Decide per person: who can create drives, view results, manage questions, run shortlisting. The team lead runs the desk; the junior runs the screen.
🎟️ Guest invitation links
Expiring invite links. No signup, no password to reset, nothing for a candidate to install. Single-purpose, time-bounded, revocable in one click.
📦 Bulk CSV operations
Five hundred candidates on one role? Drop a CSV. The platform enrols them, sends the invites, and reports back with row-level errors if anything fails.
📊 Per-client analytics
Conversion ratios, time-to-shortlist, integrity flags broken out per client, so you know which contract is your best and which one needs a conversation.
AI VOICE PRE-SCREEN
Hear the candidate before the client does.
Your recruiters can't do a 20-minute call with every CV that lands. Send a Paraakh voice link and the candidate gets interviewed role-specific questions, follow-ups when they hedge, a transcript and a fit score in your inbox. You submit the ones who actually sounded like the role.
◆
Role-aware questions
The voice agent reads the JD and asks what matters not a generic screening script.
◆
Follow-up when they hedge
Vague answer? The agent probes. You see the second answer, not the rehearsed one.
◆
Transcript + fit score
Skim a one-pager instead of replaying a call. The good ones rise to the top by themselves.
Paraakh AI
"Walk me through how you'd design the retry logic for a payment webhook."
Arjun
"I'd use exponential back-off, probably three retries…"
Paraakh AI ↳ follow-up
"What happens to the event after the third retry?"
FIT SCORE83 / 100
Defend the shortlist · proctored evidence
When the client asks "did they actually take this?" you have the recording.
Every assessment runs with face match, second-person detection, tab-switch logging, and a full audit trail. If a client questions a score, you forward the evidence pack. The mandate doesn't slip on "we're not sure who took it."
Live monitoring dashboard
Watch the whole room. Without being in the room.
Every active candidate, on one screen. Live webcam thumbnails, real-time violation alerts, and a one-click warning the candidate sees instantly for as long as the exam is running.
Live drive · Senior Frontend Engineer LIVE
89Active
78On track
9Warnings
2Critical
PS
12:04 Priya S. 88
On track OK
LW
14:32 Lin Wei 91
On track OK
DR
09:18 Diego R. 71
Tab switch ×2 WARN
AS
11:22 Ankit S. 84
On track OK
SM
15:47 Sara M. 67
Devtools detected ALERT
RP
08:55 Raj P. 79
On track OK
KT
13:09 Kiran T. 82
No face 4s WARN
AM
10:41 Arjun M. 76
On track OK
Real-time updates via Socket.io · Severity classified per violation · Every event saved to the post-exam review queue.
For Recruitment Agencies
Make every submission a strong one.
Forward us a JD and three CVs. We'll send back three scorecards. That's the whole product eight minutes, no dashboard to learn.