backend/requirements.md

1. System Requirements — MVP for AI-Powered Hiring Assessment Platform 1.1 Overview A platform for managing hiring assessments using AI, serving two primary user types: 1. HR (Human Resources) 2. Candidate Goal: Enable HR to create smart assessments, manage them, and review results easily, while allowing candidates to take assessments and review their results. 1.2 Primary Users 1.2.1 HR A user responsible for creating and managing assessments, questions, and reviewing candidate results. 1.2.2 Candidate A user who takes assessments and reviews their previous results.
2. Functional Requirements — HR 2.1 Authentication and Account Management • HR login using email and password. 2.2 Job Roles and Assessment Management • Create a new assessment. • Enter assessment information:
3. Job Title
4. Experience Level
5. Required Skills
6. Job Description
7. Number of Questions
8. Assessment Description (optional) to be displayed to the candidate 2.3 Creating Questions Using AI • Create questions through AI chat based on job information. • Save questions in a question bank linked to the assessment. • Edit questions manually. • Edit a question using AI (rewording, difficulty adjustment, clarity improvement). 2.4 Managing Question Properties • Mark a question as:
9. Failing Question / Knockout • Add tags to questions to be used as evaluation criteria (e.g., Python, Communication, SQL). • Set weight or importance for each question. 2.5 Assessment Evaluation and Review • Display assessment results in a table containing:
10. Candidate Name
11. Assessment Date
12. Time Taken to Complete
13. Final Numeric Score
14. Pass/Fail status based on:  Required level  Knockout questions
15. Candidate score per tag (Tag-based Scores) • Review a specific candidate’s answers for each question. • Display a summary view:
16. Average Scores
17. Pass Rate
18. Key Strengths and Weaknesses (from AI) • Set a score/rating for each question manually or with AI assistance. 2.6 Retrieving and Managing Assessments • Retrieve specific assessment information (questions, settings, results). • HR can edit any question and use AI for assistance. 2.7 Sharing Assessments • Generate a unique link for the assessment. • Enable/disable the assessment link.
19. Functional Requirements — Candidate 3.1 Authentication and Account Management • Candidate login (email + password). 3.2 Taking the Assessment • Access the assessment via link. • Display assessment instructions. • Answer the questions. • Submit the assessment upon completion. 3.3 Reviewing Results • Display previous assessment results. • Show overall score. • Display general feedback (optional).
20. Evaluation • Each question has a maximum score. • Questions can be marked as Knockout. • For Knockout questions:
21. Candidate must achieve ≥ 50% of the question score to pass the assessment. • Final score = sum of all question scores. 4.1 AI Evaluation • The system sends candidate answers to AI for evaluation. • AI returns:
22. Suggested score for each question.
23. Brief rationale/feedback (optional in MVP). • HR can:
24. Modify any question score.
25. Accept or ignore AI evaluation. 4.2 Language • The system interface and questions are in English only. 4.3 Single HR • Only one HR exists (no multi-company or multi-HR support in MVP).
26. MVP Core Requirements 5.1 Question Types • Text Answer questions • True / False questions • Multiple Choice questions 5.2 Evaluation • Support manual evaluation by HR. • Support AI evaluation suggestion (final decision by HR). 5.3 Permissions • HR can only see assessments they created. • Candidate can only see their own assessments.
27. MVP Assumptions • No multiple HR roles (each HR is independent). • No company/team system in the first version. • No advanced anti-cheating mechanisms (camera or tracking).