AssessMe replaces unreliable manual grading with AI-assisted scoring — while treating AI usage as a metric, not a disqualifier.
A streamlined workflow designed for engineering teams who value signal over noise.
Create questions via the UI or import a JSON file. Define your answer key, allocate marks, and set time limits. Multiple choice, free text, and code questions supported.
Register a candidate and generate a UUID access link. No account creation required. Candidates access the assessment directly — clean, distraction-free, auto-saving.
On submission, GPT-4o marks free-text and code answers against your key. You get a numeric score, qualitative feedback, and an AI usage percentage — then override anything you disagree with.
Every decision made to reduce noise and surface genuine signal.
Free-text and code answers are marked against your answer key. Each response gets a score, confidence level, and qualitative feedback — queued as background jobs so nothing blocks.
Plagiarism detection produces a percentage per answer, rolled up across the assessment. It informs your judgement — it never auto-disqualifies. A candidate who used AI responsibly shouldn't be penalised.
Syntax-highlighted code editor with language selection. Examiner sets a preferred language and can provide starter snippets in multiple languages. When a candidate is skilled, syntax is semantics.
No account registration for candidates. Generate a unique link, share it manually. The candidate confirms their name and begins. Auto-save ensures no progress is ever lost.
Set a closing deadline, a timer from first login, or no limit at all. Configure expiry behaviour per assessment — warning, grace period, or auto-submit.
Deploys behind your existing Nginx reverse proxy with zero configuration changes. All settings live in docker-compose.yml. PostgreSQL with pgvector-ready for Phase 2 AI features.
Designed for the people on both sides of the screening process.
Sign in to the examiner dashboard to create an assessment, register a candidate, and see AssessMe in action.