Feedbacker

Feedbacker is an open-source project for higher-education assessment teams designing accountable ways to use AI. It suits educators and institutions that want human approval, auditable workflows and control over deployment. The main caveat is decisive: the earlier feedback-generation application has been retired, so current value lies in the project’s rebuilding direction rather than a ready-made product.

Overview

Feedbacker is being redefined and rebuilt in the open as a workflow, governance and assessment layer around AI. That makes it a different proposition from a conventional marking assistant. The project is concerned with how assessment work is structured, reviewed and governed, with educators retaining responsibility for marks and feedback released to students.

For a university evaluating AI adoption, this emphasis is practical. It places academic judgement with staff instead of an automated decision-maker and frames AI as a controlled part of an institutional process. It is less suitable for someone seeking an immediately available service that generates feedback today, because the former application is no longer the active product.

Key features

Human approval and traceability

Feedbacker’s direction calls for a traceable chain covering inputs, criteria, transformations, suggestions, edits and approvals. In practice, that gives moderation and governance teams a framework for examining how an AI-assisted result reached its final form. Educators remain accountable for both grades and released comments, so the system is intended to support judgement rather than replace it.

Institutional control

The project treats models, prompts, data handling, retention and deployment as choices institutions should be able to govern and replace. That is a strong fit for universities with policy, privacy or procurement requirements that rule out an opaque, fixed AI service. These are stated design goals of a project under reconstruction, not assurances that every control is already delivered in a finished interface.

Rubric-led cohort work

Rubric-led workflows are intended to support moderation and calibration across cohorts while preserving academic discretion. This should appeal to departments seeking more consistent criteria without forcing identical judgements. The retired application previously described rubric-driven assessment, batch generation of criterion-referenced draft feedback, and educator control over rubrics and final decisions; those capabilities belong to the earlier product and should not be assumed to remain available.

Pricing

Feedbacker is free and open source. That removes a licence charge and allows institutions to consider how they want to host or adapt the project, but it does not establish a supported subscription, service-level commitment or a current quota structure. The former application offered public deployment or institutional self-hosting, with greater control over data and AI configuration when self-hosted; its retirement means those arrangements should not be treated as a current purchasing option.

Platforms

Feedbacker is listed for Linux, macOS, web and Windows. That broad platform coverage is useful for mixed campus environments, although the project’s rebuilding status makes the practical availability of each platform dependent on the current implementation.

Who it's for

Feedbacker is aimed at higher-education educators and institutions working with assessment and feedback. It is a good conceptual fit for teams that need educator sign-off, cohort moderation, auditable processing and the ability to govern models and data practices. It is a poor fit for an individual lecturer who needs a stable, immediately deployable feedback generator, or for buyers requiring established plans, support terms and operational guarantees.

Pros and cons

Pros

Educator authority is explicit: marks and released feedback stay with academic staff rather than an automated grading system.

Governance is built into the ambition: traceability and replaceable model, prompt, retention and deployment choices address institutional accountability.

Open-source foundation: institutions can assess the project as a basis for adaptation instead of accepting a closed workflow.

Cohort consistency is a target: rubric-led moderation and calibration can help departments align practice while retaining judgement.

Cons

The previous application is retired: readers cannot rely on its former feedback-generation and batch-processing capabilities as current functionality.

Rebuilding creates uncertainty: the present direction is clearer than the finished operational feature set.

No commercial plan structure is established: teams seeking defined support, quotas or service commitments will need another option.

Alternatives

For a broader rubric-software shortlist, see Rubric Software.

Gradescope is the better choice when you need a freemium web, Android and iOS service with a free Basic plan covering dynamic rubrics, question-by-question grading, statistics, regrade requests and export.

Choose UniRubric instead for a one-time free lecturer marking job with rubric-anchored output, approval before release and PDF export through its web or API platform.

Rubric Scorer suits a mobile workflow where the free app supports up to three rubrics for one class.

GradeBoard is an alternative for a web-based free tier with 10 learner submissions per month, limited rubric uploads, annotated DOCX download, basic analytics and 30-day retention.

OpenRubric is an alternative for self-hosted or web use under a paid model with a free first partner event.

Crowdmark is an alternative for course-based or annual assessment subscriptions across web, mobile and API platforms.

Pawtograder is an alternative for instructors wanting a free community instance with self-hosted or web deployment.

RubriqFlow is an alternative offering free unlimited faculty and staff accounts, with student access charged per student per semester.

Verdict

Choose Feedbacker if your priority is an open-source foundation for educator-led, governable AI assessment in higher education. Its strongest reason to exist is the insistence on human authority, traceability and institutional control. Look elsewhere if you need a functioning feedback-generation product, predictable commercial terms or a mature deployment today; Feedbacker’s active project is still being rebuilt.