Acai Grading Assistant helps instructors review and grade student work more efficiently and consistently. It uses AI to read student submissions and rubric information, then suggests ratings and draft feedback so you can focus on final professional judgement instead of repetitive reviewing.
Acai Grading Assistant is available in supported FeedbackFruits activities (such as Assignment Review and discussion grading in Interactive Study Materials) for institutions that have Acai enabled and the relevant feature flags turned on. To enable this feature, please contact your FeedbackFruits representative.
What is Acai Grading Assistant?
Acai Grading Assistant is an AI-powered helper for instructors during grading. It analyzes student document submissions and Acai Assisted rubric criteria to provide:
Suggested ratings at the rubric level
An explanation of how it arrived at that rating
Draft feedback comments aligned with your rubric or criteria
The goal is to streamline grading while keeping you fully in control. Acai Grading Assistant does not replace the teacher or make final grading decisions for you. Instead, it provides a starting point that you can review, adapt, or ignore.
How Acai Grading Assistant helps you
With Acai Grading Assistant, you can:
Speed up grading
Use suggested ratings and draft comments to reduce repetitive reading and writing for similar assignments.
Improve consistency
Apply criteria more consistently across students and cohorts by starting from AI-suggested ratings and explanations.
Focus on judgement, not boilerplate
Spend more time on edge cases, nuance, and pedagogy, and less on retyping similar feedback over and over.
Support detailed feedback
Use AI-generated suggestions as a base and then personalize and refine them for each student’s work.
Where you can use Acai Grading Assistant
Acai Grading Assistant is available in selected FeedbackFruits tools where instructors review student work and provide evaluations, specifically:
Assignment Review activities with Acai Assisted rubric criteria
Interactive Study Materials (ISM) with a ‘Rate students’ discussion contributions’ step and Acai Assisted criteria.
It currently works on document submissions (text-based uploads). It does not yet support non-document submissions such as standalone audio or video, and cannot read text that appears only in images within a document.
Exact availability may depend on your institution’s configuration, pilot setup, and feature flags. If you are unsure whether Acai Grading Assistant is available in a specific tool, please contact your FeedbackFruits representative.
How it works in practice
When you open a student’s submission in a supported activity with Acai Grading Assistant enabled:
Check rubric and criteria
The assistant uses Acai Assisted rubric criteria and level descriptions to understand how work is being evaluated.
Analyze the submission
It reads the student’s most recent document submission and identifies how it aligns with those criteria.
Generate suggestions
It proposes:
A suggested rating at the relevant criterion level (shown as an Acai “Suggested” chip in the rubric)
An explanation summarizing why this rating was suggested
Optional draft feedback you can adopt and edit when writing comments for the student
You stay in charge
You choose which suggested ratings or comments to accept, edit, or discard.
You decide on final feedback wording and any scores or grades.
No changes are applied without your input, and Acai’s internal suggestions do not directly affect grading or progress unless you choose to use them.
Data and privacy
Acai Grading Assistant uses advanced Large Language Models hosted on secure infrastructure (for example, Azure OpenAI Service) to process student submissions and generate suggestions.
In line with Acai principles:
Student data is processed only to generate grading assistance.
Input is not used to train, retrain, or improve the underlying language models.
We use the models in a “plug-and-play” manner with custom instructions, not by fine-tuning them with student data.
We follow FeedbackFruits’ existing security and privacy standards and regional requirements.
For more details on how data is handled, please refer to your institution’s documentation and the Transparency Note for Acai Grading Assistant and related Acai features.
Limitations and good practice
Support, not automation
Acai Grading Assistant is a support tool. It does not replace teacher judgement, institutional policies, or existing quality assurance processes.
Requires Acai Assisted criteria
The feature only works on rubric criteria that are configured as Acai Assisted. If a rubric is not “acaiable,” or if work was submitted before criteria became Acai Assisted, suggestions will not appear.
Document-based only
Acai Grading Assistant currently works on text-based documents and does not read images, audio, or video content.
Language
Today, Acai Grading Assistant is supported in English. Other languages are under consideration and may be added in future updates.
Best practice:
Keep your rubrics and criteria clear and specific so Acai can differentiate levels reliably.
Read suggestions critically and adapt them to your context and policies.
Use the assistant to handle routine patterns and repetitive comments, and invest your time where nuance and human insight matter most.