Feedbacker

What Feedbacker is

Feedbacker is an open-source higher education project being rebuilt as a workflow, governance, and assessment layer around AI for educators. Its current direction emphasizes educator responsibility: marks and released feedback remain decisions for educators, not an automated system. The project aims to make assessment steps traceable, including inputs, criteria, transformations, suggestions, edits, and approvals. It also describes institutional control over models, prompts, data handling, retention, and deployment, with those elements intended to be governable and replaceable. Planned rubric-led cohort workflows are intended to support moderation and calibration while preserving academic judgement. The previous Feedbacker application has been retired, so its former functions describe an earlier product rather than the current project. That application supported rubric-driven assessment and batch generation of criterion-referenced draft feedback; educators controlled rubrics and final judgements, and it did not automate grading decisions. Institutions could use its public deployment or self-host it, with data and AI configuration control when self-hosting. Feedbacker is described as an open-source project by Steve Huckle and lists Linux, macOS, web, and Windows platforms.