Assessment & Exam Platform

Part of the worlds-biggest-software-project initiative.

An open-source, AI-native assessment platform combining adaptive question delivery, secure exam administration, and privacy-respecting integrity monitoring.

The Assessment & Exam Platform is a candidate project to build a standards-compliant assessment system for higher education, professional certification bodies, and corporate L&D teams. It addresses the gap between expensive proprietary proctoring suites with biometric privacy concerns and barebones quiz tools that lack psychometric rigour. The platform centres on QTI-compliant item banks, adaptive testing, and LLM-assisted authoring and scoring.

Why Assessment & Exam Platform?

Incumbent proctoring vendors such as Proctorio, Honorlock, and ExamSoft rely on facial recognition and webcam surveillance, which has triggered lawsuits, GDPR challenges, and documented algorithmic bias against students with darker skin tones.
High-stakes platforms like ExamSoft (~$60/student/year) and institutional proctoring contracts ($50–200k/year) price out smaller institutions, certification bodies, and the long tail of training programmes.
The only major open-source competitor, TAO Testing, is GPL-2.0, creating strong copyleft obligations that block embedding in commercial or proprietary derivative products without a paid licence.
No incumbent offers real-time, learner-specific question generation calibrated to a demonstrated knowledge gap; static IRT-based item banks dominate even in the most advanced adaptive systems.
LLM-based scoring of short-answer and essay responses is largely absent or assistive-only; instructors still bear most of the open-format grading burden.

Key Features

Authoring & Item Banks

Multi-format question authoring: multiple choice, short answer, essay, file upload, and code items
QTI 3.0 import/export for content portability across platforms
Item bank versioning with role-based access for authors, reviewers, and administrators
Bloom's-taxonomy tagging of items for curriculum alignment
LLM-assisted question generation from learning objectives or curriculum descriptions

Secure Exam Delivery

Browser-level lockdown for timed online exams without invasive biometric capture
Randomised question and answer ordering for academic integrity
Offline-capable delivery application for high-stakes or low-connectivity contexts
LTI 1.3 integration with Canvas, Blackboard, Moodle, and D2L Brightspace
Role-based access for authors, reviewers, proctors, and administrators

Adaptive Testing & Psychometrics

Computer Adaptive Testing engine with configurable IRT-based item selection (per IMS CAT 1.0)
Per-question difficulty and discrimination analytics
Score distribution and pass/fail reporting
Hyper-adaptive item selection (backlog) using LLM-generated items calibrated to a learner's demonstrated knowledge gap

AI-Assisted Scoring & Diagnostics

LLM-based short-answer and essay scoring against instructor-defined rubrics
Per-student conceptual gap diagnostic reports synthesised from per-question performance
Plain-language post-exam diagnostic synthesis for instructors and students

Privacy-Preserving Integrity

Behavioural integrity signals fusing keystroke dynamics, typing cadence, and clipboard events
Alternative to facial-recognition-based proctoring, reducing GDPR, BIPA, and bias exposure
Transparent, explainable flagging that surfaces evidence behind each integrity alert
Optional device-switching detection for secondary screens (backlog)

AI-Native Advantage

Unlike incumbents that bolt AI on top of static IRT item banks and webcam surveillance, this platform is designed AI-native from the start. LLMs generate Bloom's-tagged items calibrated to specific learning objectives, score open-ended responses with rubric alignment at near-human accuracy, and produce per-learner conceptual gap reports that go beyond pass/fail. Behavioural signal fusion replaces invasive facial recognition with a continuous integrity signal built from keystroke, clipboard, and interaction patterns, addressing the bias and privacy criticisms that have pushed several institutions away from existing proctoring vendors.

Tech Stack & Deployment

Standards-aligned: IMS QTI 3.0, IMS CAT 1.0, LTI 1.3, IEEE 1484.12.1 LOM, WCAG 2.2, ISO/IEC 23988
Deployment modes: self-hosted (institution-controlled data residency) and managed cloud
Integrations: LTI 1.3 for major LMSs, SAML / OpenID Connect SSO, REST API for SIS roster sync
Compliance targets: FERPA, GDPR, SOC 2 Type II
Optional offline desktop delivery client for high-stakes contexts

Market Context

The global online assessment software market was valued at approximately $8.6 billion in 2025 and is projected to reach $18.4 billion by 2034 (CAGR ~8.9%), with the online proctoring sub-segment growing at ~17% CAGR toward $2.83 billion by 2031. Incumbent pricing ranges from $15–30 per proctored session and $8–60 per user per year for assessment platforms, up to $50–300 per candidate for certification testing. Primary buyers are higher education institutions, professional certification bodies (medical, legal, IT), corporate L&D teams, and government and military training programmes.

Project Status

This project is in the research and specification phase.
Contributions, feedback, and domain expertise are welcome.

Contributing

We welcome contributions from developers, domain experts, and potential users. See CONTRIBUTING.md for guidelines.

Important: All contributions must be your own original work or clearly attributed open-source material with a compatible licence. Copyright infringement and licence violations will not be tolerated and will result in immediate removal of the offending contribution. If you are unsure whether a piece of code, text, or other material is safe to contribute, open an issue and ask before submitting.

Licence

Licence to be determined. See discussion for context.