Skip to main content
An interviewer is your AI screener for one role. You set the job, pick how candidates are screened (voice, SMS, form, resume, or AI Scenario Assessment), add questions and scoring, then share a link or connect your ATS. Here’s the flow you’ll see in the app.

​
1. Start from the dashboard

From the dashboard, click Create Interviewer to start a new interview experience.
Image
You’ll go through: Generate Interview Plan → Settings (deadline, retakes, links, budget) → Interview Stages → Questions & Scoring → Budget & Notifications, then share your interview.

​
2. Generate your interview plan

First, tell HeyMilo what role you’re hiring for so it can suggest an interview plan.
  • Paste a job description — Paste your role description and HeyMilo will generate a customized interview plan.
  • Use a template — Search job roles (e.g. Account Executive, Customer Support Representative) and pick a template to start from.
  • Job posting URL (optional) — Enter a URL to pull the job description from a job posting page.

​
Compose Agent

Compose Agent is for when you want HeyMilo to generate the whole interviewer for you: which stages to use, the questions, the scoring, and how responses are evaluated.
Image
Get a full draft workflow to review and tweak:
  • Templates — Pick a role template as a starting point when it fits.
  • Add instructions — You can even say the flow you want in plain language (e.g. SMS → resume → voice).
  • Paste context — Job copy, intake notes, or a meeting transcript so the draft matches how you actually run the screen.
Always open the draft and adjust stages and questions before you go live. Pick one, then click Generate Interview Plan. → How the Composer Agent Works

​
3. Configure how it looks to candidates

Next you’ll see Configure: set how this interview appears to candidates.

​
Interview URL

  • Interview URL Name — This becomes part of the link you share (e.g. f8e-customer-support-specialist). You can keep the generated one or change it when creating the interviewer. After go-live, each candidate also gets a unique Interview URL on their profile (see Reviewing Candidates).
  • Position Title — The role title candidates see (e.g. “Customer Support Specialist”). Has a character limit.
  • Internal name (optional) — After you create the posting, set this from the Interviewers table on your dashboard if your team needs a different label (e.g. a requisition ID or client name). Used on the dashboard only; candidates still see the Position Title. Leave empty to use the public title. See Your Dashboard.
Image
Review and adjust, then click Continue to Interview Workflow.

​
4. Set up your interview stages

Here you build the sequence of screening steps. Each step is an AI agent (voice interview, resume screening, SMS, or form).
  • Available Agents — Voice Interview, Resume Screening, SMS Screening, Form Screening, AI Scenario Assessment. Click a card to add and configure that stage.
  • Your Interview Flow — Stages you add appear here. Drag to reorder how candidates move through them.
  • Templates — Use “Select a template” to quickly set up stages for common role types.
You need at least one stage before you can continue. See Interview Types for when to use voice, SMS, form, resume, or AI Scenario Assessment.

​
5. Interview settings (optional)

Before you go live you can set:
  • Agent identity — Single identity (seamless) or multiple (handoff experience).
  • Language — Default interviewer language or auto-detect candidate language (BETA).
  • Timeline — Deadline and “days to complete” after a candidate is added.
  • Retakes — Whether candidates can retake if they don’t pass.
  • Communication & branding — Email template group, design template, phone number for SMS/voice, whether to email the report to the candidate, and candidate intake fields.
  • Budget & notifications — Optional budget; email subscribers for completion alerts.

​
Candidate Intake Fields

If you want candidates to fill out specific fields before they start — like their LinkedIn profile, phone number, or earliest start date — use Candidate Intake Fields under Communication & Branding.
Candidate Intake Fields in Communication and Branding
Click Add field to library to create fields — each field has a name and a format type (File, Date, Number, Phone, URL, etc.). Once added, check the box next to any field to include it on this interviewer. Fields saved to the library are available across all your interviewers. Submitted answers appear under Registration Fields on the candidate’s profile.
Registration Fields on candidate profile
How budget enforcement works: Budget is checked when a new candidate is invited. Candidates already engaged in the flow won’t be cut off mid-experience (so their interview stays intact), which means actual spend can slightly exceed budget.Example: Import 1,000 candidates into an agent with a $10 budget → HeyMilo will engage all of them and exceed the cap, because they were already invited before it kicked in.Recommendation: Set a budget that matches the number of candidates you plan to move through the agent (per-candidate cost × planned volume + a small buffer).
You can also add a scheduling link, redirect link, and advanced settings (company info, role details, etc.).

​
Essential Settings

Under Essential Settings, configure how questions are generated for this experience:
  • Standard Generation (default) — Fast, independent per-modality generation. Each stage (voice, SMS, form, resume) generates questions separately using dedicated prompts.
  • Adaptive Generation (BETA) — Agentic generation with cross-modality awareness. Questions are generated with intelligent reuse across stages, so the overall experience feels more cohesive.

​
Timeline & expiration

While creating the interviewer: Settings → Timeline & Completion. Set when the posting expires for new candidates and how many days to complete after someone starts. Later on the dashboard: Open the interviewer and check Deadline at the top. Edit here to extend dates or time to finish.
Image
Link expired or invalid?: Almost always check the posting expiration or days to complete.

​
6. Questions & scoring

For each stage (e.g. Voice Interview), open Set Up Your Interview Questions.
  • Add questions — Your own, from the Question Library, or reuse from another interviewer. You can reorder and set how many follow-ups the AI can ask (see Follow-up questions below).
  • Evaluation criteria — For each question, define what a strong vs weak answer looks like (rubric). Set a weight so more important questions count more toward the score.
  • Enforce objective evaluation — Optional: require candidates to meet a minimum score to pass.
    Set Up Your Interview Questions with Enforce Objective Evaluation toggle and question cards showing the Max Follow-ups badge

​
Follow-up questions

On voice/video and SMS stages, the AI asks follow-up questions automatically based on what the candidate says — you don’t write each follow-up in advance. What you control is how many follow-ups the AI can ask per question, and what it probes for. Each question card shows its current cap as a Max Follow-ups badge (visible in the screenshot above). To set follow-ups for a question:
1

Open the question editor

Click the gear icon on the question card to open Edit Question.
2

Set the Follow-ups Range

Expand Advanced Criteria and adjust the Follow-ups Range slider. Min is the fewest follow-ups the AI will ask; Max is the most it’s allowed to ask (e.g. “1 follow-up”). Set the range to 0 if you don’t want any follow-ups on that question.
3

Steer what the AI probes (optional)

Use Interviewer Instructions in the same panel to guide follow-up behavior in your own words — e.g. “If the candidate mentions lack of experience, ask about transferable skills.” This guides the live conversation only; it’s not used for grading (put grading guidance in Grading Instructions instead).
4

Save

Click Save. The question card updates with the new Max Follow-ups badge.
Edit Question modal with the Follow-ups Range slider and Interviewer Instructions field under Advanced Criteria

Edit Question → Advanced Criteria: Follow-ups Range and Interviewer Instructions.

Form interviews don’t have follow-ups — forms are fixed questionnaires, so candidates only see the questions you wrote. If you need probing follow-ups, use a voice/video or SMS stage. See Interview Types.
Details: Configuring Your AI Interviewer and How Scoring Works.

​
7. Share your interview

When setup is done, you’ll see Share with Candidates:
  • Interview link — Copy and share this link so candidates can start the interview. You can also print a poster.
  • Connect to your ATS — Connect an ATS so candidates and results sync automatically. See ATS Integrations & quickstarts.
  • Ingest — Use the webhook or ingestion gateway to send candidates from job boards, your site, or other tools.
From there you can return to the dashboard, go to the agent, or set up ingestion.

​
Quick tips

  • Duplicate to save time — Duplicate an existing interviewer from the dashboard to reuse stages, questions, and settings, then adjust. You can also use the Clone an interviewer API (Next Gen reference).
  • Start simple — One voice or one SMS stage is enough to begin. Add more stages once you’re comfortable.
  • Preview — Use Preview in the top nav to see how the interview looks to candidates before you share.

​
Next steps