Skip to content
A team having a thoughtful conversation in a welcoming office

Angel AI by Arcanext

A natural conversation.A clearer next hire.

Meet Angel AI. Your first-screen phone conversations, handled. Your candidate evidence, organized. Your hiring judgment, always yours.

AI phone screening. Human hiring decisions.

  • Role-specific conversations
  • Transcript-backed evidence
  • Human-led decisions

THE CONVERSATION, CONSIDERED

A natural first hello.
A meaningful next step.

Angel follows your role-specific questions and gives candidates room to explain their experience. Explore an illustration of how that conversation becomes evidence for your team.

Illustrative demoPhone screening

Angel AI

Arcanext’s AI recruiting assistant

Space for a real answer.

An example of the conversation

Angel

“Tell me about a time you helped a customer work through a difficult issue.”

Candidate

“I walked through the issue with them, then wrote a guide the team could reuse.”

Your questions, in a natural conversation
Conversation becomes evidence.Transcript · Structured answers · Rubric scores
Illustrative interface and fictional conversation. No live call or audio.

The Angel workflow

From first hello
to an informed next step.

Let Angel do the repetitive part. Keep your team close to what matters.

  1. 01

    You set the direction

    Your role. Your questions.

    Define the role, write your interview questions, and set a weighted scoring rubric. Upload the candidates you want to screen.

    A first screen shaped by your team

  2. 02

    Angel handles the conversation

    A phone call, not another form.

    Angel calls candidates, introduces itself as AI, and works through your questions with contextual follow-ups. The conversation is recorded and transcribed.

    Consistent questions. Room for context.

  3. 03

    You make the call

    Review the evidence. Decide.

    Explore structured answers, rubric scores, and supporting transcript quotes. Revisit the recording, then choose who moves forward.

    The hiring decision stays with you

Explore the Arcanext workflow
Screening evidenceSample scorecard

Customer support specialist

Fictional candidate · Example rubric

Problem solving

Rubric weight · 50%

4 / 5

“I walked through the issue with them, then wrote a guide the team could reuse.”

Supporting transcript excerpt · Sample 02:14

Cross-team collaboration

Rubric weight · 50%

3 / 5

“I shared the guide with my team. I haven’t worked with the engineering team directly yet.”

Supporting transcript excerpt · Sample 03:08

Context for your review

A concrete support example. Cross-functional experience needs a closer conversation.

Sample weighted score

70 / 100

Human decisionPending review
Fictional scores and excerpts, for illustration only. An AI assessment is not a hiring decision.

The words behind the score

Don’t just see a score.
See what shaped it.

A number is a starting point, not the whole person. Angel brings the conversation into the evaluation, so you can look beyond the summary.

  • 01

    Your rubric, made visible

    Review criterion-level scores against the standards and weights your team defined.

  • 02

    Evidence you can revisit

    Read supporting quotes, explore structured answers, and return to the transcript and recording.

  • 03

    Context, not just a conclusion

    Use strengths, concerns, and the original conversation to guide your own follow-up.

Built to support your judgment. Not substitute for it.

Candidate-first, by design

A person on the other end.
Never just a profile.

The first conversation sets the tone. Make it clear, considerate, and grounded in the role.

A clear introduction
Hi, I’m Angel, an AI recruiting assistant calling about the customer support role. This call is recorded for the hiring team to review.
Illustrative opening · Not a live call
01

Clear from the first hello.

Angel identifies itself as an AI recruiting assistant, explains the purpose of the call, and discloses recording before interview questions begin.

02

Their local time matters.

Outbound calls follow country-specific calling windows in the candidate’s local time. Outside the window, the call waits.

03

A missed call is not a rejection.

Candidates who don’t answer remain in the No-Answer lane. Their hiring decision stays pending, not automatically rejected.

A deliberate boundary

Angel screens.
People decide.

Automation can organize a conversation.
It shouldn’t have the final word on a person.

Angel never automatically hires or rejects a candidate. AI can make mistakes, and consistent criteria do not guarantee bias-free outcomes. Review the evidence, question the assessment, and bring your own judgment.

  1. AI evidence
  2. Human review
  3. Your decision

A little more clarity

Good questions.
Straight answers.

What Angel does, what it doesn’t, and where your team stays in control.

Explore pricing
What does Angel AI actually do?

Angel AI is Arcanext’s AI phone-screening product. Your team defines a role, interview questions, and a scoring rubric, then uploads candidates. Angel conducts the phone screen and returns a recording, transcript, structured answers, and rubric-based evaluation for your team to review.

Can we use our own questions and scoring criteria?

Yes. Hiring managers author the interview questions, choose which are required, and define criteria with rubric descriptions and weights. Angel uses those questions to guide the conversation, with contextual follow-ups bounded by the question’s follow-up policy.

Do candidates know they’re speaking with AI?

Yes. The call opens with an introduction identifying the assistant as AI, the reason for the call, and a recording disclosure. This happens before interview questions begin. Calls also respect country-specific calling windows in the candidate’s local time.

Does Angel AI decide who gets hired or rejected?

No. Angel provides evidence and an AI assessment, not a final decision. Shortlisting, rejection, and hiring decisions belong to a human. Screening status is kept separate from the hiring decision, and a missed call never triggers an automatic rejection.

Does a consistent rubric make screening bias-free?

No. A shared rubric can make the evaluation structure more consistent, but it does not guarantee fairness or eliminate bias. AI can make mistakes, and the questions and criteria people choose also matter. Review the original evidence, consider missing context, and keep people accountable for decisions.

Does Angel AI integrate with our ATS or calendar?

Not in the current product scope. Angel focuses on uploaded candidates, phone screening, and evidence review. ATS integrations, calendar scheduling, candidate sourcing, and email, SMS, or WhatsApp outreach are not part of V1.

Meet your new first-screen partner

A better first conversation
starts with Angel.

Let Angel handle the phone screen.
Keep the next hiring decision in your hands.