Hi Anaya, a senior Go role on a payments team. Hybrid, with the salary band in the brief.
Anaya
Interesting. How many days in the office?
Zia
Three days in Bangalore. I can send you a link to a voice interview.
Zia
Sent: your interview link. Start any time, or pick a time that suits you.
Naira report cardVoice · Naira
Maya Thompson·Applied for Collections Specialist·Interviewed Sep 27, 2026, 38 min
74.0/ 100 overall
AlignedIntegrity 96%English (CEFR) B2
Strongly not alignedAlignedStrongly aligned
“Clear, calm caller who secured a payment plan without pressure tactics; audit caught 4 of 5 compliance slips. Speaks fluently at a natural call pace.”
Handle a Customer Who Disputes a Past-Due Balance
M2Voice only
Module Score4.1 / 5
Strengths
Verified identity before discussing the balance
Offered a concrete 3-month plan unprompted
Weaknesses
Slight hesitation stating the exact due date
Did not repeat the total balance for confirmation
Evidence (transcript quotes)
“I understand this is frustrating — let's find a plan that works for you and confirm it today.”
Naira report cardVoice · Naira
Maya Thompson·Applied for Collections Specialist·Interviewed Sep 27, 2026, 38 min
74.0/ 100 overall
AlignedIntegrity 96%English (CEFR) B2
Strongly not alignedAlignedStrongly aligned
“Clear, calm caller who secured a payment plan without pressure tactics; audit caught 4 of 5 compliance slips. Speaks fluently at a natural call pace.”
Handle a Customer Who Disputes a Past-Due Balance
M2Voice only
Module Score4.1 / 5
Strengths
Verified identity before discussing the balance
Offered a concrete 3-month plan unprompted
Weaknesses
Slight hesitation stating the exact due date
Did not repeat the total balance for confirmation
Evidence (transcript quotes)
“I understand this is frustrating — let's find a plan that works for you and confirm it today.”
An AI recruiting agent sources, contacts, screens and interviews candidates on its own, with a human approving key steps. How it works, what it cannot do, and a worked example.
An AI recruiting agent is software that pursues a hiring outcome on its own: it sources candidates, contacts them, screens and interviews them, and builds a shortlist, choosing its steps as results come back. A human approves the steps that spend money or contact people and makes the hiring decision. A tool or chatbot does one step when asked.
AI recruiting agent vs AI recruiting tool vs chatbot
The phrase now covers a search box, a scheduling bot and a system that runs a whole hiring process. The difference that matters is not how clever the model is. It is how much of the work happens without a person, and what happens at the points where a person is still required.
What it does
How you can tell
Chatbot
Answers candidate questions and collects details in one conversation.
It reacts to whoever writes to it. It never starts work on a role.
AI recruiting tool
Does one step well: searches profiles, drafts a message, schedules, or scores a resume. You act on every output.
Nothing happens if you close the tab.
Automation
Runs a fixed sequence you configured. Same steps, same order, every time.
It cannot decide to do something different, only to stop.
AI recruiting agent
Pursues a hiring outcome. Chooses its steps, reacts to what comes back, and stops at gates a person controls.
Two runs of the same brief can legitimately look different.
An applicant tracking system (ATS) is a different thing again: it is the record of who applied and where they are. An agent does work; an ATS stores the result of work. Most agents either sit on top of an ATS or hand their shortlist to one.
If two runs of the same brief could never differ, it is automation. That is fine. It is just not an agent.
What an AI recruiting agent does, step by step
Source. Turns a job description into a searchable profile, searches for people who match, and ranks them with a reason for each score. A human should confirm the profile before the search spends anything.
Reach out. Writes to each candidate, follows up on a schedule, and stops the moment someone replies or opts out. A human should decide whether messages go out on their own or wait for approval.
Screen.Answers the candidate’s questions about the role and checks the facts that rule people in or out: location, notice period, the must-have requirements.
Interview. Runs a structured first interview, the same plan and rubric for everyone on the role, and writes a report with the evidence behind each score.
Shortlist. Hands over the people worth meeting, with the reasons. The hiring decision stays with a person.
Products in the category split this loop differently. Paradox’s Olivia screens and schedules applicants by chat and text. LinkedIn’s Hiring Assistant runs searches across LinkedIn, drafts outreach and pre-screens interested candidates. Juicebox describes agents that source around the clock and personalised outreach sequences. HireVue pairs assessments with an AI interviewer that runs voice-based interviews. Those descriptions come from each vendor’s own site, read on 3 November 2026; products change, so check before you decide. A side-by-side of the field is in best AI recruiting agents in 2026.
Example: one agent per job at EmployLabs
EmployLabs splits the loop into three agents that share one role, so each job has an owner you can inspect.
Meera sources. She reads a job description or one sentence, asks every missing question in one batch, and builds a role brief you confirm. She searches professional-profile data providers in an order an administrator sets, moving to the next provider only when the earlier one has nothing for the search, and ranks every match with the reason behind its score. A requirement the profile cannot answer is marked unknown, never counted as a fail. You can see a sample of real profiles before anyone is contacted.
Zia engages. She reaches candidates on email, WhatsApp, SMS, calls, X and Reddit, answers their questions about the role, and when a candidate is interested and the role allows it, sends the interview invite. A reply stops the follow-ups.
Naira interviews.She runs a live voice interview from a browser link, in English, with modules chosen by the role’s seniority. A live interview runs up to 60 minutes, each module up to 30. Where a role needs it, she sets a take-home with a follow-up call where the candidate defends the work, or a skills test. Every score in the interview report cites the candidate’s own words.
Where humans stay in the loop
Approvals. Nothing runs until a person arms the role and chooses how much runs on its own. Approving a send or a spend is a human click; the agents cannot approve their own work or raise their own autonomy. That rule is enforced in code rather than requested in a prompt.
Overrides. You can change a requirement and the pool is scored again against it, step outreach down, lower a cap or disarm the role at any point.
Final decisions. The agent recommends; a person decides who is interviewed by the team, who gets an offer and who is hired. At EmployLabs an offer is recorded for the pipeline; the agents do not send it.
What hiring metrics improve with automation?
The metrics that typically improve are the ones made of waiting: time to the first shortlist, how fast a candidate hears back, how many candidates reply, recruiter hours per hire, and the share of interviews that end in an offer, because the people reaching the hiring team were screened against the same bar. Quality of hire is slower to read and depends on the decision a person makes. Measure each against your own baseline before you switch anything on; a vendor’s average is not your number. We cover which measures stay honest once software does the work in recruiting metrics that survive automation.
Risks and rules
Bias. An agent that ranks people can rank unfairly at scale. Ask how scores are explained, and whether every candidate on a role is measured against the same rubric.
New York City Local Law 144. Employers using an automated employment decision tool for candidates in New York City need an independent bias audit and must tell candidates the tool is used.
Ask to see the reasoning behind one decision it made.
Go through the candidate experience yourself, including declining.
Ask what stops it spending, and what a refusal costs.
Ask what you keep if you stop paying.
Common questions
What is an AI recruiting agent?
An AI recruiting agent is software that pursues a hiring outcome on its own: it sources candidates, contacts them, screens and interviews them and builds a shortlist, choosing its steps as results come back. A human approves the steps that spend money or contact people, and makes the hiring decision.
How is an AI recruiting agent different from an ATS or a chatbot?
An ATS is the record of who applied and where they are; it stores the result of work. A chatbot answers whoever writes to it. An agent starts and carries the work itself, from finding people to interviewing them, and stops at gates a person controls.
Can an AI agent run the whole hiring process?
An agent can run most of the hours: sourcing, outreach, follow-up, screening and a structured first interview. The decisions stay with people: who the team meets, who gets an offer and who is hired. A well-built agent also needs a person to approve what it spends and who it contacts.
Do AI recruiting agents replace recruiters?
No. They take over the high-volume, time-sensitive work, so each recruiter can carry more roles. Persuading a hesitant candidate, pushing back on an unrealistic brief and choosing between two good finalists still need a person.
What hiring metrics typically improve with automation?
Typically time to the first shortlist, time for a candidate to hear back, reply rates, recruiter hours per hire and the share of interviews that end in an offer. Measure each against your own baseline before switching anything on.