Candidate report view
Review the full AI assessment for an individual candidate, switch between answers and analytics, and capture follow-up actions.
Page overview
Clicking any candidate card from the pipeline opens their detailed report. The top section shows the candidate name, status badge, job title, and seniority level. Three tabs below let you switch between different views of their interview results.
Candidate report header
Example showing Michael - Building Architect position
Michael
Tabs
Two tabs organize the candidate information: Overview shows scores, strengths, and key metrics, while Answers displays all interview responses with search and filtering options. The page remembers which tab you're viewing and keeps your filters active when switching between them.
Overall evaluation
The Overview tab starts with a summary card showing the AI recommendation, final score (displayed as a circular badge), and five skill ratings. Each rating uses a 0–10 scale for Technical, Communication, Soft Skills, Cultural Fit, and Writing Quality.
Scores come from the AI analysis of all interview answers. The system looks at answer quality, depth of knowledge, and how well the candidate communicates their experience. Higher scores indicate stronger performance in that area.
Michael - Building Architect (Overall evaluation)
Final recommendation with circular score and skill breakdown.
Progress metrics
Below the skill scores, you'll see a metrics grid showing how many questions were answered, completion percentage, typing pace (questions per minute), and total interview duration. These metrics help you understand the candidate's engagement level and time management.
Michael - Building Architect (Progress metrics)
Completion status and interview pacing details.
96%
Answered 24 of 25 questions.24
Session completed early at question 24.0.8
Fast pace — verify answer depth.Question limits
The system tracks how many questions the candidate was expected to answer. If they finish early, the progress percentage adjusts automatically to reflect their actual completion.
Strengths & weaknesses
Two side-by-side cards display the candidate's main strengths and weaknesses, pulled from the AI evaluation. Each card has a toggle button that expands or collapses the detailed bullet list, letting you quickly review key patterns in the candidate's answers.
Michael - Building Architect (Strengths)
Expandable card showing key competencies.
• Leads stakeholder workshops confidently
• Connects research insights to design outcomes
• Strong technical problem-solving skills
Candidate profile sidebar
The right side of the Overview tab shows two information cards: Candidate Profile and Job Summary. The Profile card displays when the interview started and ended, plus any contact links the candidate provided (email, GitHub, LinkedIn, resume, website).
Michael - Building Architect (Profile & job summary)
Interview dates and contact information.
Building Architect
Seniority: Senior • Contract: Full-timeScore interpretation & red flags
Use these guidelines to understand what AI scores mean for hiring decisions. Always combine the numerical ratings with a review of the actual answers for a complete assessment.
Final score ranges
Score interpretation tiers
Recommended action thresholds based on thousands of evaluations.
Strong match — Advance to phone screen immediately
Good fit — Review answers tab, schedule interview if skills align
Borderline — Check weaknesses, consider junior roles or contract work
Likely mismatch — Archive unless specific strengths compensate
Context matters
Score ranges shift based on seniority and role complexity. Junior positions may advance candidates scoring 65+, while senior/lead roles typically require 80+ for interviews. Always review the Strengths & Weaknesses cards before final decisions—a 72 with perfect technical skills but weak soft skills may outperform an 81 with generic strengths.
Red flags to watch for
Beyond AI scores, watch for these warning signs in candidate reports that may indicate poor fit or interview quality issues.
Critical red flags 🚩
Automatic disqualifiers in most cases
❌ Copy-pasted / generic answers
Responses feel templated, lack specifics, or identical phrasing across questions❌ Contradictory statements
Claims 5 years React experience but can't explain basic hooks❌ Refusal to answer key questions
Skips technical questions or responds with 'I don't know' to fundamental topics❌ Extremely low QPM (0.2)
Took 2+ hours for 20-minute interview, suggests disengagement or cheating❌ Suspiciously high QPM (1.0)
Answering too fast with perfect responses may indicate AI assistance (ChatGPT, etc.). Cross-check answer quality vs speed.
Moderate red flags ⚠️
Investigate further before rejecting
⚠️ One-word or very short answers
Most responses under 10 words, suggests lack of effort or poor communication⚠️ Off-topic responses
Answers don't address the question asked, AI may have scored low accurately⚠️ No concrete examples
Claims experience but provides zero specific projects, tools, or outcomes⚠️ Incomplete interview
Answered only 40% of questions then stopped (check if technical issue or disinterest)⚠️ High technical + low communication scores
Check behavioral answers for articulation ability - may struggle in team environments
Green flags for strong candidates
These positive indicators suggest a candidate will excel beyond what the AI score alone reveals.
Excellence indicators ✅
Signs of top-tier candidates
✅ Specific, detailed examples
Describes exact projects, tools used, metrics achieved (e.g., 'reduced load time by 40% using Redis caching')✅ Proactive problem-solving
Answers show they anticipated challenges and took initiative without being asked✅ Continuous learning mindset
Mentions recent courses, side projects, or self-taught skills beyond job requirements✅ Clear communication
Complex technical answers explained simply, logical flow from problem → solution → outcome✅ Cultural alignment
Work style, values, and team preferences match your company culture (e.g., prefers async communication for remote role)
When to override AI scores
AI scoring is highly accurate but not perfect. Use human judgment to override in these specific scenarios.
Override upward (give candidate another chance)
When to interview despite low AI score
Scenario 1: Misunderstood question format
Candidate gave correct answer but in unexpected format (e.g., wrote code snippet when AI expected explanation)
Scenario 2: Rare specialized skill trumps other gaps
Score is 65 overall but candidate has critical niche expertise (e.g., mainframe modernization) that's impossible to find
Scenario 3: Strong soft skills compensate for technical gaps
Junior role where trainability > current knowledge. Candidate shows exceptional communication and learning agility.
Override downward (reject despite high AI score)
When to reject even with 85+ score
Scenario 1: Detected plagiarism/cheating
Answers are technically perfect but clearly copied from Stack Overflow (exact matches found via Google search)
Scenario 2: Cultural misalignment
High technical score but work style preferences fundamentally incompatible (e.g., wants 100% async in office-first company)
Scenario 3: Overqualified + flight risk
Candidate is VP-level applying for mid-level role. Likely using as stopgap and will leave quickly.
Answers tab controls
The Answers tab shows a search bar and question type filter at the top. You can search across both questions and answers (the search updates as you type), and the dropdown lets you filter by question type: Behavioral, Technical, Situational, or Follow-up.
Michael - Building Architect (Answer filters)
Search and filter controls.
Answer details
Each question and answer pair appears in an expandable accordion. The collapsed view shows the question text, a preview of the answer, question type badge, timestamp, and response time (color-coded by speed). Expanding the accordion reveals the full answer text plus detailed evaluation scores and specific strengths or weaknesses identified by the AI.
Michael - Building Architect (Answer accordion)
Expandable question with answer preview and metadata.
Tell us about a project where you collaborated with engineers.
"We ran weekly syncs with backend to align on schema changes..."
Expanded question with full evaluation
When you expand a question accordion, you see the complete answer text plus a detailed evaluation panel. The evaluation shows three skill scores (Technical, Communication, Soft Skills) as progress bars, followed by specific strengths and weaknesses identified by the AI.
Michael - Expanded answer with evaluation
Full answer text with detailed AI scores and feedback.
I would start by establishing clear communication channels — for example, using shared platforms like BIM 360 or Slack for daily updates and coordination. I'd schedule regular virtual design reviews to ensure consistency and maintain the project's design intent.
Follow-up questions: Adaptive intelligence
The system automatically generates follow-up questions when the AI detects an insufficient answer or very low scores. This adaptive approach gives candidates a second chance to demonstrate their knowledge and provides deeper insights into their capabilities.
Follow-up questions appear visually indented with a lightning icon (⚡) and a vertical guideline connecting them to their parent question. The badge shows "Follow-up" in blue to make the relationship clear.
Michael - Follow-up question
Indented follow-up with lightning icon showing adaptive questioning.
Can you describe a specific instance where you resolved a coordination conflict between disciplines in Revit, and what steps did you take to address it?
On a mixed-use development project, we discovered that several HVAC ducts were clashing with structural beams in the Revit model...
Why follow-ups matter
Follow-up questions serve three critical purposes: (1) They give candidates who were nervous or misunderstood the question a fair second chance, (2) They provide deeper technical validation by asking for specific examples when initial answers were too vague, and (3) They help identify candidates using AI tools—generic first answers followed by equally generic follow-ups suggest copy-pasting rather than genuine knowledge. The AI considers both the original and follow-up answers when calculating final scores.
Exports & downloads
The overflow menu in the header exposes a “Download raw JSON” action. It streams the current payload to a file named , preserving answers, reports, and the job metadata for offline review.
Download menu
Three-dot menu with JSON export option.
Browser support
The export feature works across all modern desktop browsers. Mobile devices trigger downloads according to the operating system's default file handling.
Next steps
After reviewing individual candidates, explore the templates library to standardise new job designs and interview flows.
Continue to Templates library