Technical Evaluation Platform
Enterprise Suite
2026
An AI-assisted technical interview app with two sides: candidates register to answer dynamic LLM-generated questions, while employers create shareable assessment links and review candidate insights from a secure dashboard. Built with session integrity tracking and multi-language support.
Snapshots
For Candidates
Dynamic LLM Questions
LLM-generated prompts tailored to stack and level, with a static question bank as fallback.
Scored Feedback
Each submission is scored and given written feedback by the model, followed by a comprehensive summary.
Skip Logic
Candidates can move on without answering; the question is stored as 'Not answered' with a zero score.
Session Integrity Tracking
Tab/window focus leaves are recorded per question to ensure assessment integrity.
Voice Input & Multi-Language
Optional browser speech recognition and full UI support for English, Tamil, Hindi, and French.
Returning Candidates
Returning emails are recognized and visit history is tracked across sessions.
For Employers
Secure Auth
scrypt-hashed passwords and HMAC-signed session cookies with middleware-protected routes.
Named Assessment Links
Create an assessment with a name and get a shareable link that automatically stamps candidate submissions.
Insights Dashboard
Drill down through three columns: technology → assessments → candidates.
Candidate Detail Popups
Click a candidate to see every submission, score, and piece of feedback in a convenient popup.
Candidate Journey
Onboarding
Candidate completes intro form (language, stack, level).
Generation
API generates targeted questions via LLM.
Execution
Candidate answers or skips; tab switches are recorded.
Completion
System runs a final LLM pass for a debrief and summary.
Employer Journey
Authentication
Employer logs in via secure middleware-protected route.
Creation
Generates a named assessment and receives a shareable link.
Distribution
Candidates use the link; submissions are tracked automatically.
Review
Employer reviews insights and candidate submissions via dashboard.
App & Routing
Next.js App Router orchestration handling public candidate forms and middleware-protected employer dashboards.
State & i18n
Zustand manages session state globally, while i18next provides seamless multi-language support.
Database & Auth
Supabase PostgreSQL handles candidates, submissions, and session integrity tracking, with custom HMAC-signed session cookies for employers.
LLM Inference
Hugging Face Inference (Qwen2.5-Coder / Llama-3.2) generates contextual questions and evaluates answers, with NVIDIA NIM fallback.
The database uses Supabase PostgreSQL with RLS policies allowing public read/insert for the interview flow.
technologiesSelectable stacks (seeded).candidatesRegistration records and total score.submissionsAnswers, scores, feedback, and assessment_id.tab_switchesFocus-leave events for session integrity.employersHiring-side accounts with unique email and password hash.assessmentsNamed assessments owned by an employer.