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.

​

How scoring works

​

How Scoring Works

​

Structured Questions vs Scored Questions

• Use a 1–5 scoring scale
• Require explicit criteria for score 1 (lowest) and score 5 (highest)
• The AI infers intermediate scores (2, 3, 4) from that continuum
• Best for open-ended, nuanced evaluation (skills, experience, communication)

• Use binary pass/fail evaluation
• Include Yes/No, Minimum Amount, and Multiple Choice types
• Can be marked as dealbreakers to automatically disqualify candidates
• Best for objective, verifiable requirements (work authorization, certifications, availability)

​

Overview of the Scoring Mechanism

​

Role of the AI in Determining Scores

• The candidate’s responses (audio, video, or text) are transcribed into text where applicable.
• This transcript, together with your defined scoring benchmarks, is evaluated by HeyMilo’s AI (a fine-tuned language model).
• The AI interprets the candidate’s response against the criteria your team defined. It uses natural language understanding to apply qualitative descriptors (e.g. “excellent,” “vague,” “significant”) in context, in line with how human evaluators would judge responses.
• The AI assigns scores by comparing the candidate’s response to your clearly defined 1 and 5 criteria.

​

How Intermediate Scores (2, 3, and 4) Are Assigned

• Intermediate scores are inferred by the AI based on the continuum you set between 1 and 5.
• The AI evaluates how closely a candidate’s response aligns with either end of that spectrum.
• Scores 2, 3, and 4 are not explicitly defined by recruiters. That keeps evaluation flexible and avoids overly rigid scoring, while still using your 1 and 5 as anchors.

​

Human Oversight and Score Override

• Recruiters have full visibility into the rationale behind each AI-generated score, including access to the full transcript.
• Recruiters can manually override any AI-generated score if they disagree or have context the AI didn’t capture.
• The final decision on candidate progression always stays with your recruiters, not the AI.

​

Mitigation of Bias

• The scoring system is transparent and auditable. Your team defines evaluation criteria in plain language, which reduces ambiguity and subjective interpretation.
• HeyMilo undergoes third-party bias audits to help ensure the scoring system remains fair and unbiased.
• Ongoing monitoring and the option for human overrides further support fairness.

​

Data Use & Confidentiality

• Candidate data and scoring information are kept strictly confidential.
• HeyMilo complies with GDPR and SOC 2 requirements. View our compliance dashboard for more.

​

What’s in a Score Report

• Overall score — Weighted average of your question scores (1–5). Weights are set by you.
• Highlights & lowlights — Quick strengths and concerns from the AI.
• Per-question breakdown — Response, score, and reasoning for each question.
• Tags — Auto-extracted (e.g. skills, experience, availability). See Interview Tags.
• Transcript — Full text; voice/video includes audio.
• Cheat detection (if enabled) — Integrity flags. See Interview Integrity.

​

Next Steps

• Reviewing Candidates — Read score reports and take action
• Structured Questions — Pass/fail and dealbreakers
• Interview Evaluation — Tune rubrics and weights

​

Scoring system overview

​

How Our Scoring System Works

Learn how HeyMilo’s AI scores candidates objectively and consistently at scale.

​

Structured Questions vs Scored Questions

• Use 1-5 scoring scale
• Require explicit criteria for scores 1 and 5
• AI infers intermediate scores (2, 3, 4)
• Best for open-ended, nuanced evaluation

• Use binary pass/fail evaluation
• Include Yes/No, Minimum Amount, and Multiple Choice types
• Can be marked as dealbreakers to automatically disqualify candidates
• Best for objective, verifiable requirements

​

Overview of Scoring Mechanism:

​

Role of the AI in Determining Scores

• The candidate’s responses (audio/video) are transcribed into text.
• This transcript, combined with the explicitly defined scoring benchmarks, is evaluated by our fine tuned Large Language Model (LLM).
  • The LLM leverages extensive natural language processing (NLP) training on vast, diverse text datasets. This allows it to interpret adjectives and qualitative descriptors (e.g., “excellent,” “vague,” “significant”) in context, similar to human evaluators. This ensures the interpretation aligns closely with general human judgement and understanding.
• Furthermore, the AI assigns scores by interpreting the candidate’s response against the clearly defined criteria reviewed by your recruiting team.

​

Data Use & Confidentiality

• Candidate data and scoring information remain strictly confidential.
• HeyMilo complies with both GDPR and SOC2 requirements. View our compliance dashboard.

​

Additional resources