Mercu leverages AI to score candidate answers to specific question types, streamlining the shortlisting process.

This scoring system enables you to prioritise candidates based on the quality of their answers without having to read, listen to, or watch hundreds of responses individually.

Instead, the AI-generated scores act as a guide, directing your focus to the strongest applications and helping you efficiently allocate your time.

Mercu supports AI scoring for the following question types:

Questions with AI answer scoring display a score (Likert scale: very weak to very strong) across the table view, kanban view, and candidate profile UI.

Hovering over the score reveals the candidate's original response—whether text, video, or voice—enabling you to quickly dive into the details beyond the AI score. This provides both an at-a-glance assessment and the option to explore each answer in depth.

Here’s an example of how we present a short answer with AI scoring: the AI score (in this case: "Very Strong") appears directly in the table cell. Hovering over it reveals both the candidate’s original answer and an explanation of the score, offering quick insight into answer quality and the reasoning behind the AI evaluation. 👇🏽

Here’s an example of how we present a video interview with AI scoring: the AI score (in this case: "Very Strong") appears directly in the table cell. Hovering over it reveals the video itself, a transcript of the video, and the rational for the AI score.👇🏽

AI answer scoring comes pre-enabled for voice and video interview modules, so there's no need for separate activation.

Additionally, our AI automatically assesses short-answer responses to ensure relevance; if a reply doesn’t meet the question criteria, candidates are prompted to resubmit. This feature is ready to use out-of-the-box, streamlining the review process from day one.

AI scoring (a 1-5 score based on the response quality) can be optionally enabled for each short-answer question.

Simply toggle on 'AI Scoring' to activate it for any future a to that question; note that past answers will not be scored retroactively.

The 'Is there a correct answer?' setting is designed for factual questions with a single, objectively correct answer. It’s ideal for questions like, "What's the only protein we offer on our menu?" but not suitable for questions like, "Do you have a food hygiene certificate?" where answers may vary.