BRIGHTSPACE · LTI 1.3
AI Essay Grader for Brightspace D2L: Grading, Feedback and Oral Exams via LTI 1.3
Connect Eduface to Brightspace D2L over LTI 1.3 to grade essays, give rubric-aligned feedback, mark open-ended exams and run AI oral exams. A lecturer approves every grade before it reaches the Gradebook.
By Eduface · June 2026 · 8 min read
Every term, learning technologists at Brightspace institutions receive the same question from academic colleagues: "Can we get AI to help with marking?" Brightspace is a mature, well-designed platform, but its built-in tools do not grade submissions or generate rubric-aligned feedback. Finding AI assessment tools that integrate cleanly, respect existing workflows, and meet data protection requirements has taken real effort. Eduface was built to fill exactly that gap.
What AI assessment tools work with Brightspace D2L?
Eduface provides four AI assessment tools that connect to Brightspace D2L via LTI 1.3: a Paper Grader for essays and assignments, a Feedback Tool for personalised formative feedback, an Exam Grader for open-ended written exams, and an Oral Examination tool. All tools pass grades back to the Brightspace Gradebook after explicit lecturer approval. Eduface is GDPR-compliant, EU AI Act compliant, and an approved supplier on the Jisc/CHEST framework for UK institutions.
How does Eduface integrate with Brightspace via LTI 1.3?
LTI 1.3 (Learning Tools Interoperability) is the current standard for connecting external tools to a VLE. It handles authentication, data exchange, and grade passback through a secure, standards-based protocol that Brightspace fully supports. For the institution, this means no custom code and no student-facing disruption: students continue to submit work through Brightspace exactly as they do today.
Integration is configured once by the institution's learning technologist. Eduface's team provides a dedicated setup guide and onboarding support. Once connected, submissions flow from Brightspace into Eduface automatically when a lecturer launches a grading session. After the lecturer reviews and approves grades, those grades pass directly back to the Brightspace Gradebook. No manual CSV export or import is needed.
For UK institutions
Eduface is an approved supplier on the Jisc/CHEST framework, which simplifies procurement. Irish institutions can access Eduface via the HEAnet framework. Bath Spa University, which runs Brightspace, is among Eduface's pilot partners.
The diagram below illustrates the full workflow from student submission to final grade.
Student
Submits via Brightspace
Eduface
AI processes submission
Lecturer
Reviews and approves grade
Brightspace
Grade recorded in Gradebook
Student submits in Brightspace, LTI 1.3 sends the submission to Eduface, the AI processes it, the lecturer reviews and approves the grade, and it returns automatically to the Brightspace Gradebook.
What does the AI Paper Grader add to Brightspace assignments?
Brightspace handles assignment submission well. What it cannot do is read a 3,000-word essay against a rubric and produce a scored, annotated draft for the marker to review. The Eduface Paper Grader does exactly that.
A lecturer uploads their rubric and assignment brief once at the start of the module. When submissions arrive through Brightspace, Eduface annotates each one: highlighting relevant passages, assigning a score to each rubric criterion, and providing written reasoning for every mark. The suggested grade is held as a draft. Nothing reaches the student until the lecturer explicitly approves it.
Lecturers can work in two modes. In blind mode, the AI suggestion is hidden: the lecturer marks independently, and Eduface's output is available as a comparison afterwards. In AI-visible mode, the AI suggestion appears alongside the submission, and the lecturer can edit any part of it before approving. In UK pilots, lecturers changed an average of just 5% of each final grade, indicating 95% alignment between AI suggestion and final marker decision.
"In UK pilots, lecturers changed an average of just 5% of each final grade."
Eduface Paper Grader pilot data
How does AI formative feedback work inside Brightspace?
Formative feedback is one of the highest-impact levers available to lecturers, and also one of the most time-intensive. Eduface's Feedback Tool generates personalised, rubric-aligned feedback for every submission. The AI drafts the feedback; the lecturer can review and edit it before it is released to the student. No student receives feedback that has not passed through a human review step.
Four feedback styles are available, and lecturers can select the one that fits their pedagogy:
• Reflective and Socratic, for courses that prioritise critical thinking
• Constructive and Direct, for vocational or practice-based programmes
• Went Well and Needs Improvement, for structured progression feedback
• Supportive and Encouraging, for early-year or widening-participation cohorts
The tool tracks student skill progression across multiple drafts, which makes it particularly useful for modules with iterative submission structures. Six domain-specific submodels are available: Law, Economics, Social Sciences, STEM, Humanities, and Health Sciences. The submodels are trained on disciplinary conventions, so the feedback reflects the norms of the field rather than generic writing advice.
How does the Exam Grader handle open-ended questions in Brightspace?
Multiple-choice questions in Brightspace grade themselves. Open-ended written exam questions do not. The Eduface Exam Grader fills that gap.
For each submitted answer, three independent AI agents evaluate the response against the grading scheme without seeing each other's assessments. A fourth agent then reconciles the three scores. Where the three agents are in strong agreement, the system flags high confidence. Where there is significant disagreement between agents, the submission is flagged for closer lecturer attention. The suggested grade, together with per-criterion reasoning, is presented to the lecturer. The lecturer fills in the final score and confirms before any grade is recorded in Brightspace.
The multi-agent approach addresses one of the central challenges in exam marking: consistency. Across the same cohort with multiple human markers, Eduface's Exam Grader has been found to be 48% more consistent than unaided human grading. Processing time is under four minutes per assignment from upload to suggested grade.
Full audit trail included
Every submission retains a complete record of the AI's per-agent assessments, the reconciliation result, and the lecturer's final decision. This is required for EU AI Act compliance and supports academic appeals processes.
How does Oral Examination support academic integrity in Brightspace courses?
Eduface's Oral Examination tool is currently in early access with pilot institutions. Lecturers configure a character and a scenario: for example, a strict dissertation examiner, a sceptical investor reviewing a business plan, or a distressed patient in a simulated clinical encounter. The AI conducts a structured oral conversation with the student and provides consistent, rubric-aligned evaluation afterwards.
The tool serves two distinct purposes. For formal oral assessment, it gives every student access to a consistent, well-calibrated examiner regardless of assessor availability. For academic integrity verification, it provides a way to test whether submitted written work genuinely reflects the student's own understanding: a student who did not write their own essay will typically struggle to discuss it under oral questioning.
Six academic fields are supported. Institutions can request early access at eduface.me/product/oral-examination.
Does Brightspace already have AI assessment tools? What are Intelligent Agents?
Brightspace includes a built-in feature called Intelligent Agents. It is a useful tool, but it is not an AI assessment tool. Intelligent Agents is an automation and notification system: it monitors learner activity against rules you define (for example, "student has not logged in for five days" or "student has not submitted by the deadline") and sends automated alerts to learners or instructors when those conditions are met.
Intelligent Agents cannot read a submission, apply a rubric, score an essay, or generate personalised feedback. It is a workflow automation tool, not an AI grading tool. Eduface is what fills the genuine AI assessment gap in Brightspace.
Summary of the distinction
Brightspace Intelligent Agents send automated notifications based on activity rules. Eduface grades submissions, generates feedback, and returns scores to the Brightspace Gradebook. They serve entirely different functions.
How does D2L Lumi compare to Eduface for AI essay grading?
D2L Lumi is Brightspace's own AI product, and a genuinely useful one for what it is designed to do. Lumi helps lecturers create course materials faster: it generates quiz questions from existing content, produces module summaries, suggests assignment and discussion ideas, and provides students with an in-platform chatbot for learning support. These are valuable capabilities for course design and student engagement.
Lumi is not an assessment tool. It does not read submitted student work. When an essay, open-ended exam answer, or written assignment arrives in Brightspace, Lumi offers no mechanism to apply a rubric, score each criterion, or generate feedback grounded in what that particular student actually wrote. There is no grade-passback workflow from an AI assessment step, no per-submission audit trail of rubric scores, and no human-in-the-loop approval step for individual student grades, because Lumi was not built for that purpose.
Lumi includes a feature that allows lecturers to build banks of reusable feedback text. This is an authoring tool: it helps a lecturer set up feedback templates in advance. Eduface does something different. It reads every individual submission and generates feedback specific to what that student wrote, criterion by criterion, at any cohort size. The 200th submission receives the same level of personalised detail as the first.
The two products are complementary. A Brightspace institution might use Lumi to streamline course design and provide learner support, and Eduface to handle the assessment workload when submissions arrive. The table below sets out where each product operates.
Capability
D2L Lumi
Eduface
Generate quiz questions from course content
Yes
Not applicable
Student learning chatbot inside Brightspace
Yes
Not applicable
Create module summaries and course materials
Yes
Not applicable
Read and assess submitted student essays
Not available
Yes
Apply a lecturer's rubric to individual submissions
Not available
Yes
Per-criterion scoring with written reasoning
Not available
Yes
Personalised feedback based on each student's submission
Not available
Yes
Grade open-ended written exam answers
Not available
Yes
AI-conducted oral examination
Not available
Yes (early access)
Automatic grade passback to Brightspace Gradebook
Not available
Yes
Human-in-the-loop approval for every grade
Not applicable
Yes
Jisc/CHEST framework approved supplier
No
Yes
EU AI Act compliant for high-risk AI assessment
Not applicable
Yes
Tool
Purpose
Assessment type
Lecturer approval point
Grade passback to Brightspace
Paper Grader
Grade essays and academic papers at scale
Written coursework
Before any grade is released to the student
Yes, automatically after approval
Feedback Tool
Personalised rubric-aligned formative feedback
Draft submissions and formative tasks
Before feedback reaches the student
Feedback only (no summative grade)
Exam Grader
Grade open-ended written exam questions
Written examinations
Lecturer confirms final score before recording
Yes, automatically after confirmation
Oral Examination
AI-conducted oral assessment and integrity verification
Oral exams and viva-style assessments
Rubric-aligned evaluation reviewed by lecturer
Yes, after lecturer review (early access)
Frequently Asked Questions About AI Assessment in Brightspace
Does D2L Lumi grade student essays or exam answers?
No. D2L Lumi is a course creation and learner support tool. It generates quiz questions, module summaries, and assignment ideas, and provides students with an in-platform chatbot. It does not read submitted student work, apply a rubric to individual submissions, or return grades to the Brightspace Gradebook. For AI grading and rubric-aligned feedback on actual student submissions, institutions connect Eduface to Brightspace via LTI 1.3.
Does Brightspace have a built-in AI grader?
No. Brightspace D2L does not include a built-in AI grading tool. The platform's "Intelligent Agents" feature automates notifications and rule-based actions, but it cannot assess written work, apply a rubric, or generate feedback. Institutions that want AI grading inside Brightspace need to connect a specialist tool via LTI 1.3. Eduface is purpose-built for this integration.
Does Brightspace's Intelligent Agents feature do the same thing as Eduface?
No. Brightspace Intelligent Agents is a notification and workflow automation tool. It tracks learner activity against rules you set and sends alerts when those rules are triggered. It has no ability to read assignment text, evaluate it against a rubric, or produce a grade or feedback. Eduface is a separate AI assessment platform that integrates with Brightspace via LTI 1.3 to provide actual grading and feedback capabilities.
Is Eduface GDPR-compliant for Brightspace users?
Yes. Eduface runs on proprietary GPU infrastructure located in the Netherlands, meaning student data never leaves the EU. Eduface does not use third-party AI APIs such as OpenAI, so submission data is not shared with external providers. The platform is both GDPR-compliant and EU AI Act compliant. Education grading is classified as high-risk AI under Annex III of the EU AI Act, and Eduface is built to meet those requirements, including full audit trails and mandatory lecturer approval before any grade is recorded.
How is Eduface procured for Brightspace institutions in the UK?
UK higher education institutions can procure Eduface through the Jisc/CHEST framework, which eliminates the need for a separate full tender process. Irish institutions can access Eduface via HEAnet. For institutions outside these frameworks, Eduface's team supports standard procurement processes. Bath Spa University, a Brightspace institution, is among the pilot partners.
Do students need a separate Eduface account when submitting through Brightspace?
No. Students submit their work through Brightspace exactly as they do today. The LTI 1.3 integration handles authentication and data transfer in the background. Students do not need to visit the Eduface platform, create an account, or change any part of their submission process. The experience is entirely transparent to them.
Connect Eduface to your Brightspace instance
Eduface integrates with Brightspace via LTI 1.3 in a single setup session. All four tools are available: Paper Grader, Feedback Tool, Exam Grader, and Oral Examination. Book a demo to see the workflow live, or create a free lecturer account to explore the platform.