Bias-aware hiring scorecard assistant for structured interviews and evidence-based ratings
A custom GPT for hiring managers and interviewers: builds a role scorecard with job-related competencies and behavioral anchors, writes structured interview questions, and turns interview notes into evidence-linked ratings while flagging non-job-related or biased language.
Agent in Operations by AISkillHub Library
What it does
Unstructured interviews and free-text feedback make it easy for impressions to stand in for evidence. Structured interviews with defined competencies and rating anchors are a widely recommended way to make hiring decisions more consistent. This GPT helps you run that process without extra paperwork.
**What it does**
Three modes:
1. **Build a scorecard.** From a job description, propose four to six job-related competencies, each with a definition and a 1 to 4 rating scale with behavioral anchors. 2. **Write interview questions.** For each competency, behavioral ("Tell me about a time...") and situational questions, plus follow-up probes. Every candidate gets the same core questions. 3. **Score interview notes.** Paste your notes for one candidate. It maps evidence to each competency, quotes the note that supports each rating, marks competencies with no evidence as "not assessed", and flags language in your notes that is not job-related (for example comments on age, accent, appearance, family, "culture fit" without definition).
**When to use it**
- Opening a new role and agreeing what "good" looks like before interviews - Writing feedback straight after an interview - Preparing for a debrief so discussion starts from evidence
**What it will not do**
- It does not make hire or no-hire decisions, rank candidates against each other, or produce a single overall score. People make the decision. - It does not assess candidates from CVs, photos, social media or names. - It does not give legal advice on employment law in your country. If your process raises a legal question, check with HR or counsel.
**Example**
Note: "Strong communicator, seemed young for the role. Walked through migrating 40 services to a new CI system, cut build time but couldn't say by how much."
Output: Technical delivery 3 (evidence: led CI migration of 40 services), Impact measurement 2 (evidence: could not quantify build time change), Communication "not assessed" (the note gives an impression, not evidence), and a flag on "seemed young": it refers to perceived age, which is not job-related, so remove it from the feedback.
How to use it
**Set up as a ChatGPT custom GPT** (requires a ChatGPT plan that lets you create GPTs).
1. In ChatGPT, open the GPTs area from the sidebar (Explore GPTs) and choose **Create**. 2. Switch to the **Configure** tab instead of using the chat-based builder. 3. Set **Name** to `Hiring Scorecard Assistant` and write a one-line **Description**. 4. Open `assistants/hiring-scorecard-assistant/instructions.md`, copy everything, and paste it into the **Instructions** field. The file is written to fit inside the field's character limit. 5. Add **Conversation starters**: `Build a scorecard from this job description:`, `Write structured interview questions for these competencies:`, `Score my interview notes against the scorecard:`, `Review my feedback for non-job-related language`. 6. Under **Knowledge**, upload (optional) your company's competency framework, career ladder or interview guidelines. The GPT will use your competency names instead of its own. 7. Under **Capabilities**: turn off web search and image generation. These are not needed and keep the GPT focused on your inputs. 8. Test it in the preview pane on the right with a real example, then save and choose who can access it (only you, anyone with the link, or your workspace, depending on your plan).
**Privacy note:** avoid pasting candidate contact details or protected information. Refer to candidates by initials or a candidate ID. Check your company's rules on putting candidate information into AI tools before use.
Files
Save each file at the path shown, relative to your project root.