An AI hiring portal for employers.

Challenge

Where hiring teams were

Those employers post a role, receive a flood of speculative applications, then hand the sorting to a keyword filter that cannot see working style or aspiration. Good candidates are screened out for the wrong words, and the ones who make it through arrive with a CV that hides as much as it shows.

Approach

What the portal does

Matched candidates, ranked by fit

The employer opens a role and finds candidates the model has already matched to it, ranked by fit, each one arriving as a person with the history and the skills behind the match visible on the card.

The reasoning behind the match

The match is broken down criterion by criterion, showing where the candidate and the role line up and where they do not, and the employer reads that reasoning before deciding whether to signal interest.

Behavioural fit alongside the skills

A behavioural read taken during candidate onboarding puts working style and team posture next to the experience, so the shortlist speaks to how someone works as well as to what they have done.

Contact gated by mutual interest

The employer and the candidate can only talk once both have opted in, so first contact is warm and informed.

The screens

Inside the employer portal

Candidates arrive already matched to the role, each with a match strength, their last two roles and the skills behind the match. The employer reviews a shortlist of people, each with the evidence for their place on it.

The candidate record carries the match insights panel, where role experience, location, responsibilities, industry and work arrangement are ticked off one by one and the product and sales-channel criteria show plainly where the candidate does not fit.

The employer writes the role once and sets the criteria the matching runs against, with the form warning in plain terms that criteria drawn too tight will return few matches or none.

Job setup is split into basics, detail, compensation and role specifics, with progress tracked against each so a half-finished role is obvious before it goes live.

The screens are the live employer portal, and the candidate and company details in them are invented.

Result

What the employer gets

Read the fit up front

The employer sees where a candidate aligns with the role and where they do not before spending an hour in an interview to find out.

A shortlist about people

Behavioural fit and working style sit next to the experience, so the shortlist reflects who the person is and how they will work with the team already in place.

Contact worth having

Mutual interest means every first message lands with a candidate who has already chosen to talk, so outreach stops being a cold-send.

A brand that does not ghost

Status is visible to both sides, so candidates are kept informed and the employer reputation is not quietly eroded by silence.

Status

Where the portal is now

The portal is live, and the employers using Chapta hire through it, from writing the role to the first conversation with a candidate.

Runs on

What it runs on

• Explainable matching model
• Behavioural assessment layer
• Mutual-interest workflow
• Employer job and candidate portal
• Real-time status tracking