Conversational ATS: The Future of Hiring Is a Two-Way Dialogue

Overview

• A conversational ATS replaces the career portal with a two-way chat, usually on WhatsApp, and writes every answer straight into the applicant tracking system.
• It fits South Africa because around 94% of the country's internet users are already on WhatsApp (MyBroadband).
• In graylink's setup, txtHR runs the conversation and Neptune scores, routes, logs and reports on it.
• Candidates get instant replies and status updates. Recruiters get a pre-screened shortlist and a complete audit trail.
• Our position: keep the candidate-facing chat rule-based, and put the AI behind it where a recruiter reviews its output.

What is a conversational ATS?

A conversational ATS is an applicant tracking system that collects applications through a chat instead of a web form. The candidate answers short questions in a messaging app, and each answer lands in the ATS as structured data.

The difference is that the system replies. It confirms the application, explains what happens next and sends an update each time the candidate's status changes.

That removes the two things candidates dislike most about portal applications: the effort of getting in, and the silence after they submit.

Why career portals lose South African candidates

Career portals assume a laptop, a stable connection, a password and a CV saved as a file. Many high-volume applicants in South Africa apply from a phone and have none of the other three.

What they do have is WhatsApp. DataReportal counted 51.7 million internet users in South Africa at the end of 2025, and MyBroadband reports that around 94% of them are on WhatsApp. Local networks also sell low-cost, WhatsApp-only data bundles.

That last detail is easy to miss. A candidate on a WhatsApp-only bundle can chat for next to nothing, but may need to buy extra data just to load a career site. Each extra step between the job ad and the first question is where non-desk applicants drop out.

Our earlier piece on WhatsApp recruiting covers the channel itself in more detail.

How txtHR and Neptune work together

txtHR runs the conversation and Neptune keeps the record. The candidate only ever sees the chat, while the recruiter works from Neptune.

This is what a typical volume-hiring application looks like, step by step:

Three practical details matter here. txtHR also works in a mobile browser, so nobody needs to download an app. It can run in the candidate's own language. And answers are checked on the phone before sending and again on the server, so recruiters are not cleaning bad data later.

Where AI is used, it sits in Neptune's RecruitAI layer, where recruiters review its output, rather than in the chat itself.

What recruiters get back

Recruiters stop sorting applications and start making decisions. Acknowledgements, knockout screening and status updates run without them, so the first thing they open is a shortlist of candidates who already meet the minimum criteria.

The admin saving shows up quickly. The City of Cape Town's HR team reports that automating everyday HR processes with txtHR reduced its admin volume and improved turnaround times.

Compliance also gets simpler. Every consent prompt, message and outcome is logged with a time stamp, so POPIA and Employment Equity reporting draws on one record instead of a spreadsheet rebuilt at month-end. The same data feeds Neptune's recruitment reporting.

None of this replaces the recruiter. It returns the hours spent chasing, sorting and re-keying, which is the case we make in why AI will not replace recruiters.

Where conversational hiring fits, and where it does not

Our view: a conversational ATS should be the default for high-volume and entry-level hiring, and a support tool for everything else.

It earns its place in retail, call centres, learnerships and the YES programme, where thousands of mobile-first applicants compete for similar roles and speed decides who you keep. For senior or specialist roles, use the chat for logistics such as interview scheduling and document collection. Screening at that level belongs with a recruiter, because strong senior candidates expect a person.

We also think the candidate-facing chat should stay rule-based, which is how txtHR is built. Every applicant gets the same questions and the same knockout rules as the person before them, and that consistency is what makes a screening decision defensible if it is ever challenged.

Generative AI belongs behind the chat, where a recruiter checks its output. We unpack that trade-off in recruitment chatbots vs AI agents, and our recruitment and AI guide covers the wider picture.

Frequently asked questions

Do candidates need to download an app to apply through a conversational ATS?

No. txtHR runs on WhatsApp, which most South African candidates already use every day, and it also works in a mobile browser for anyone who does not have WhatsApp.

Is a WhatsApp job application POPIA compliant?

It can be, provided consent is captured before any personal information is collected and every message is stored with an audit trail. txtHR asks for consent in the chat, and Neptune keeps the time-stamped record for reporting.

Can a conversational ATS handle thousands of applicants for one vacancy?

Yes. The chat runs around the clock and screens every applicant against the same knockout questions at once, so volume only reaches the recruiter at the shortlist stage.