Feedback Tasks
Create tasks, share them with learners, and let the AI grade them automatically – with detailed feedback for each criterion.
How feedback tasks work
With Feedback Tasks you create assignments with clear assessment criteria. Learners submit their answers and instantly receive AI-powered feedback – per criterion with points and an explanation. No login required for learners.
Create
Create a task manually or let the AI generate it
Share
Distribute a link or QR code to learners
Submit
Learners write their answer – no login needed
Feedback
AI grades automatically based on your criteria
Why feedback tasks with sevenAI?
Automated, criteria-based feedback – developed specifically for education.
Instant feedback
Learners receive detailed feedback within seconds – not days or weeks later. This enables faster learning progress.
Criteria-based, not generic
Unlike ChatGPT, sevenAI does not give generic feedback but evaluates each of your self-defined criteria individually – with points and specific feedback per criterion.
No login for learners
Share tasks via link or QR code. Students do not need their own account – no registration hassle, no data privacy concerns.
GDPR compliant
All data stored on European servers. Student texts are not used for AI training – a crucial difference from ChatGPT and other US services.
Multiple submissions
Learners can revise and resubmit their answer. This turns the task into a learning process – not just a one-time test.
Track results at a glance
View all submissions per access link at a glance. Export individual assessments as PDF – ready for documentation.
sevenAI vs. ChatGPT for feedback
sevenAI Feedback Tasks
- ✓Individual assessment criteria with points
- ✓Adjustable grading strictness
- ✓Access link without student login
- ✓PDF export of the assessment
- ✓GDPR compliant, European servers
- ✓Consistent, reproducible grading
General AI tools (ChatGPT etc.)
- ✗Unstructured free-text feedback
- ✗No consistent grading scale
- ✗Each student needs their own account
- ✗No export, only copy-paste
- ✗US servers, data usable for training
- ✗Results vary for the same input
Create task
Create manually
Define all details yourself:
- ●Title of the task
- ●Task description – detailed instructions
- ●Model answer – as a reference for AI grading
- ●Assessment criteria – as many as you like, with individual point values
- ●Grading strictness (Lenient, Standard, Strict)
- ●Max. submissions per learner
AI-generated
RecommendedProvide just a few parameters – the AI creates everything:
- ●Topic – e.g. “Argumentative essay on school uniforms”
- ●Grade level – e.g. 10th grade
- ●Difficulty – Easy to Challenging
- ●Text type – e.g. argumentative essay, summary, essay
- ●Language – Optional, default is German
The AI automatically generates: title, task description, complete model answer, and matching assessment criteria with points. You can adjust everything afterwards.
Understanding assessment criteria
You define how many criteria your task has and how many points each criterion is worth. The AI grades each criterion individually. Here is an example with 5 criteria:
Share & Access
Create access links
For each task you can create as many access links as you need – e.g. one per class or course group. Learners do not need their own account.
Direct link
Share via email or messenger
QR code
Print on a worksheet or project on screen
Expiry date
Optionally set a time limit
What learners see
Learners open the link and see:
- ✓The task description in detail
- ✓Attached documents to read
- ✓The assessment criteria with maximum point values
- ✓A text field to enter their answer
- ✓After submission: Instant detailed feedback
AI Assessment
How does the AI grade?
The AI acts like an experienced teacher: it compares the submission with the model answer (as a reference, not as the only valid template) and grades each of your criteria individually.
Analyse task description & criteria
The AI understands the context, requirements, and grading standards.
Review submission
Each criterion is checked individually and scored (0 to max.).
Formulate feedback
A constructive comment per criterion plus an overall summary feedback.
Grading strictness
Lenient
Full points if the content is essentially correct. Minor errors are tolerated. Ideal for practice phases.
Standard
Balanced grading. Minor deductions for inaccuracies. The recommended mode for most tasks.
Strict
Precise requirements. Language and formal errors are taken into account. Suitable for exam preparation.
Good to know
- ●The model answer serves as a reference, not as the only valid answer. Alternative approaches are accepted.
- ●Identical submissions are automatically detected – the AI does not grade them twice.
- ●Learners can submit multiple times (configurable) to improve.
Results & Export
View submissions
- ✓All submissions per access link at a glance
- ✓Detailed view: submitted answer, total score, feedback per criterion
- ✓Status overview: Pending, Being graded, Graded
Export
- PDF export – Formatted document with task, answer, and assessment
- Text export – Plain text for quick copying and further processing
Example: What the feedback looks like
Overall feedback
“Overall a well-written argumentative essay with a clear structure. The argumentation is coherent but could be strengthened with more concrete examples. Linguistically confident, few formal errors.”
All key arguments are present.
Clear organisation, logical structure.
Appropriate style, few repetitions.
More concrete examples would be desirable.
Only occasional spelling errors.
Create your first feedback task
From task creation to automatic feedback in just a few minutes.
Get started