Skip to content
Splashgain

The everyday problem

HR spends its week on formatting.

Offer letters retyped from the last one, CTC arithmetic done in a spreadsheet, four hundred CVs read by whoever has the time, and the same policy question answered again. None of it needs a specialist. All of it needs to be right.

What it does

Six agents, one role

Letter generation agent

An offer letter in minutes, with the arithmetic right.

  • Offer, appointment, experience and relieving, revision, promotion
  • CTC structure computed rather than typed
  • Your letterhead, your clauses, your signatory

Shortlisting agent

A ranked shortlist with the reason beside every name.

  • Scores every applicant against the job description
  • Criteria and weights are yours, and visible
  • Ranks and explains — it never rejects

AI interview agent

Every shortlisted candidate interviewed the same way, this week.

  • Questions drawn from the CV and the job description
  • Video, on the candidate’s own time, in their language
  • Scored against your rubric, with the answer beside the score
  • A person reviews the recording and decides

Policy answering agent

HR stops answering the same leave question.

  • Answers from your own handbook, quoting the clause
  • Says it does not know rather than improvising
  • Available where your people already are

Onboarding and exit agent

Nobody discovers on day one that nothing was raised.

  • Runs the checklist across IT, finance and the manager
  • Chases the step that has not been done
  • Confirms asset return and access removal on exit

Compliance calendar

The statutory date arrives with the work prepared.

  • Tracks the filings and renewals you are subject to
  • Prepares the paperwork ahead of the date
  • Escalates before a deadline, not after

Ready to see it in action?

Send one job description and a stack of CVs — we’ll send back a ranked shortlist with the reason beside every name.

Or call +91 95525 86428

What changes

Before and after

TodayWith the agent
Offer lettersRetyped from the lastGenerated, then signed
CTC arithmeticA spreadsheet, by handComputed every time
Four hundred CVsWhoever has timeRanked, with reasons
Policy questionsAsked of a personAnswered from the handbook

How the work moves

Who does which step

  1. 1Your team

    Request

    A hire, an exit, a revision.

  2. 2AI agent

    Prepared

    Drafted on your template, figures computed.

  3. 3AI agent

    Checked

    Against policy, dates and past letters.

  4. 4Your team

    Signed

    A person signs. Always.

  5. 5AI agent

    Filed

    Stored, indexed and logged.

Where it stops No letter is issued and no candidate is rejected without a person. The agent prepares and ranks; a human signs and decides.

What it prepares

Five letters, one signature each

What it computesWho signs
Offer & appointmentFull CTC structureDirector
Experience & relievingTenure and last working dayDirector
Salary revisionOld and new structureDirector
PromotionGrade, title and revisionDirector
Internship extensionRevised end dateDirector

500+

Customers worldwide

17+
Countries served
35+
AI agents running in our own production
100%
Platform uptime

Questions we are asked

Before you commit

Does it decide who gets hired?

No. It ranks applicants against criteria you set and shows the reason for every position, then a human reads the list. It has no ability to reject anyone, and the candidates it ranks lowest are still in the file for you to look at.

What about bias in shortlisting?

It scores against criteria you define and weight, and shows its reasoning for each candidate — so a criterion that is producing a biased result is visible and can be argued with. It does not learn a preference from your past hiring, which is the mechanism by which most automated screening reproduces bias.

Where does employee data sit?

In your systems, on your infrastructure or a cloud region you choose. Salary and personal data are exactly the category we would expect a client to insist stays in-house, and the architecture assumes that from the start.

Can it use our existing letter formats?

It uses nothing else. Your templates, your clauses, your letterhead and your signatory. We start from the letters you already issue rather than a generic set.

Does it integrate with our HRMS?

It connects to what you run rather than replacing it. Where an HRMS has no usable interface, it works from the exports you already produce — which is slower to set up but not a blocker.

Can it answer employees directly?

Yes, on the channel your people already use, and only from your own handbook with the clause quoted. Anything it cannot answer from the handbook goes to a person rather than being improvised.

Try it on one role.

One job description, one stack of CVs, one ranked shortlist. Judge it against what your team would have produced.

Book an Agent Readiness Sprint

Or call +91 95525 86428

Book a demo

See it running on your process