How the agent talks to candidates
The path of the WhatsApp conversation: opening, questions, evaluation, outcome, language, skipped questions and handover to a person.
Updated on Oct 12, 2026
The agent asks the candidate the job role's questions on WhatsApp, evaluates them and tells them the next step.
Where to find it: conversations are in Inbox. The agent's settings are in Agent.
The path of the conversation
- Opening: the candidate receives the Opening Template. With Automatic pre-screening on in the campaign, the conversation starts as soon as the candidate applies.
- Questions: the agent asks the questions of the job role or campaign.
- Answers to the candidate's questions: it uses the Agent, job role and campaign FAQs and its general knowledge.
- Evaluation: it applies filters and criteria and calculates the score.
- Outcome: the candidate is a fit or not a fit. A candidate who is a fit moves on, for example with the interview invitation.
Good to know
- Transparency: all templates have the footer Conversation handled by an AI assistant.
- Hours: the agent replies 24 hours a day, holidays included.
- Language: the conversation starts in the Campaign language (the campaign's Setup block). Then the agent keeps the language the candidate writes in.
- Style: depends on the Empathy Level. With None the agent asks the questions exactly as they are written.
- Skipped questions: the agent skips a question only if it's sure the candidate has already answered it.
- Unclear answers: the agent asks for clarification at most once, then moves on.
- Stop communications: the candidate can pause and then resume WhatsApp messages with the button in the opening template.
- Buttons: in Inbox you also see the WhatsApp buttons shown to the candidate.
- Candidate in several campaigns: they are contacted in chronological order. See How a candidate in several campaigns works.
When the agent hands the chat to a person
The agent stops and opens a human intervention request when a rule is triggered:
- Request to speak with a human: the candidate asks for a person (you also get a browser notification);
- Protected categories: the candidate says they belong to a protected category;
- Aggressive or frustrated tone;
- a rule you created.
See What a human intervention request is and how it differs from the Inbox and How to set up human intervention rules.
Common issues
The agent skips questions. It only does this if it's sure the candidate has already given the answer, even earlier in the chat. If the answer isn't there, leave feedback on the conversation.
The agent speaks the wrong language. Check the Campaign language in the campaign's Setup block: it's the starting language. After that the agent follows the candidate's language.
The agent asked for a person for no clear reason. Open the request: the message that triggered it has a label with the reason. Often the rule is too generic.
The candidate no longer gets messages. They may have used Stop communications. Check the chat in Inbox.
Related articles
- What happens if the candidate doesn't reply (follow-up and 24-hour window)The automatic follow-up after 6 hours, the expired WhatsApp chat and how to contact a candidate again with a template.
- What the Agent is and what you can configureThe Agent section sets the WhatsApp profile, language, empathy, templates, FAQs and rules of the AI assistant that talks to candidates.
- How to set up the Agent's FAQsEnter the answers the agent gives to candidates' most common questions, valid in all campaigns; what general knowledge is.
- How to customize interview outcome and closing messagesEdit the messages the agent sends after the interview (fit, not a fit, to hire) and when you manually mark a candidate as not a fit.
- How to edit the agent's WhatsApp templatesEdit the messages the agent sends with a template: first contact, follow-up, interview outcomes and manual rejection. Meta approves each template within 48 hours.
- Why did the agent discard a candidate?How to understand why a candidate turned out not to be a fit (Evaluation tab, filters, criteria, answers) and how to step in.
- Why doesn't the evaluation match the chat?The evaluation is based on the conversation and a few details from the CV, like the date of birth. What to check when the outcome seems different from what you read in the chat, and how to correct it.
- How to set up human intervention rulesDecide when the agent stops the conversation and hands it to a person; default and custom rules.