Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.heymilo.ai/llms.txt

Use this file to discover all available pages before exploring further.

​
Structured Questions

They are not scored on the 1–5 scale. Outcomes are pass or fail only. They do not change the candidate’s overall numerical score. For nuanced answers that should rank candidates, use scored questions — see How Scoring Works.
Remember: Structured = binary screening. Scored = judgment + contribution to overall score. Use both in the same interview when it fits.

​
Structured vs scored (quick compare)

StructuredScored
OutcomePass / fail1–5
Overall scoreDoes not contributeContributes (with weights)
DealbreakerOptional auto-failNo
Best forVerifiable requirementsBehavior, depth, communication
Has 1–5 rubric / weightNoYes

​
When to use structured questions

Good fits
  • Legal / policy: work authorization, licenses, certifications
  • Numeric bars: minimum years with a tool, hours available, counts
  • Knockouts: relocation, schedule, physical requirements, clearance
  • Knowledge checks with a clear correct answer (multiple choice)
Poor fits
  • Open-ended or behavioral prompts (“tell me about a time…”)
  • Anything that needs nuance or grading on a spectrum → use scored questions

​
Question types

​
Yes / No

You set: expected answer (Yes or No). Dealbreaker (optional): fail anyone who answers wrong. Examples
  • Work authorization — expect Yes, dealbreaker on
  • Physical requirement (e.g. lift 50 lbs) — expect Yes, dealbreaker on
  • “Available to start within two weeks?” — expect Yes, dealbreaker off if it’s a preference
  • “Willing to work occasional evenings/weekends?” — expect Yes, dealbreaker on if non-negotiable
Tips: One requirement per question. No double negatives. Phrase for clarity when read aloud by the AI; consider non-native speakers.

​
Minimum amount

You set: minimum value, optional unit (e.g. years, hours per week, certifications). The candidate passes if their answer meets or exceeds your minimum. Dealbreaker optional. Examples
  • Warehouse experience — min 3, unit years, dealbreaker on
  • Python experience — min 3, unit years, dealbreaker on
  • Weekly availability — min 20, unit hours per week, dealbreaker off
  • Safety certifications — min 2, dealbreaker on
Tips: Name what you’re counting clearly. Keep thresholds realistic (e.g. don’t ask for impossible years on a new technology). The AI may probe equivalents in follow-ups — still keep the question itself precise.

​
Multiple choice

You set: up to four options, optional extra context per option, which option(s) count as correct (one or more). Dealbreaker optional. Examples
  • Shift willingness — options for morning / afternoon / evening / weekends; mark the combinations you require
  • Equipment certified to operate — mark required certs (e.g. forklift + pallet jack)
  • Languages or stacks used professionally — mark must-haves for the role
  • “Comfortable with which databases?” — mark required vs nice-to-have; dealbreaker off if flexible
Tips: Cap at four options. Use short context to disambiguate. Don’t mark an unrealistic number of options as “correct.”

​
How to add them

When creating or editing voice/video questions:
  1. Add New Question
  2. Open the Structured tab (not scored)
  3. Choose Yes/No, Minimum Amount, or Multiple Choice
  4. Enter question text
  5. Configure type-specific fields (expected answer, minimum + unit, options + correct answers)
  6. Turn on Mark as dealbreaker only if a fail should disqualify
  7. Save
Structured questions do not use 1–5 criteria or scoring weights — only pass/fail configuration.

​
Mixing structured and scored in one interview

Example flow
  1. Structured Yes/No (dealbreaker): authorized to work in the US
  2. Structured minimum (dealbreaker): years of required skill
  3. Scored: challenging project, depth of experience
  4. Scored: judgment / communication
  5. Structured multiple choice (no dealbreaker): nice-to-have tool exposure
Structured items filter and label pass/fail. Scored items rank via the overall score. That combo gives knockouts plus fair comparison on the rest.

​
Best practices

  • Put dealbreakers early so you don’t spend depth on candidates who can’t meet basics.
  • Wording: specific, unambiguous, one ask per question; test how it sounds out loud.
  • Dealbreakers: only for true must-haves — too many shrinks your funnel.
  • Pilot: run the interview yourself; have a teammate try it; check pronunciation of technical terms.
  • Multiple choice: use additional context sparingly but clearly — don’t bury candidates in text.

​
Common mistakes

  • Unrealistic minimums (e.g. more years than the tech has existed) — align with the role and market.
  • Compound Yes/No (“cloud and Docker?”) — split into two questions.
  • Too many “correct” MC options — define what “qualified” actually means.
  • Structured for subjective traits (“strong leader?” Yes/No) — use a scored behavioral question instead.

​
Reviewing results

For each structured item you’ll typically see:
  • Pass / fail
  • What the candidate answered
  • Short evaluation explaining the outcome
  • Tie-in to the transcript where helpful
Structured outcomes sit beside the overall score — they don’t move the 1–5 aggregate. Failed dealbreakers should surface in your workflow for quick filtering and review.
Tip: Even without dealbreakers, structured pass/fail fields are easy scan points when comparing candidates — still with no impact on the numerical overall score.

​
Next steps

Need something that isn’t covered here? Contact support@heymilo.ai.