Hugo is TimelyGrader's AI Assistant. He works alongside you inside the platform, helping you build rubrics from scratch, personalize feedback, and make adjustments post grading.
Hugo is not agentic.
We could have made Hugo agentic. Feed him the rubric and submissions, let him give you final grades and feedback. We decided early on that was not the intent for Hugo, partly because...
When we spoke with our first ten institutional partners, eight of them were strongly against making Hugo agentic, especially for assessment. Faculty and admins told us that they wanted AI that helped them, not make decisions on their behalf.
"In the early days of planning and working with our institutional partners, we decided to not make Hugo agentic, meaning that he will help but he won't do it for you."
Hugo can create, update, or refine rubrics from scratch. If a criterion is producing inconsistent scores, he can help you diagnose why and suggest revisions.
Hugo helps you set up assignments and settings so TimelyGrader grades the way you want it to: feedback structure, grading instructions, inline comment settings, citation formats, and Content Bank uploads.
When reviewing a submission, Hugo can clarify the rationale behind a suggested score, pinpoint where a student was unclear, or help interpret a criterion in an edge case. He can also refine feedback tone, shorten comments, or translate feedback into different languages.
COMING SOON...
Hugo can apply feedback and grading changes across all submissions at once, so a mid-batch adjustment does not mean going submission by submission.
On Data and Training for Hugo...
When your department chair or IT asks "does this tool learn from our students' work?", the answer is no.
Most AI tools learn from how you use them. Hugo does not work that way.
Hugo uses assignment context, your rubric, the student submission, to help you grade better and provide better feedback in the moment. That context is not used to train any model. Your institution's grading data stays on your dedicated TimelyGrader instance and does not feed back into Hugo's underlying intelligence.
Hugo Help You Learn!
What Hugo does instead is try to make you better at using the platform. If you are unsure how a criterion applies to an edge case, Hugo explains the reasoning. If your rubric is producing inconsistent results, Hugo helps you diagnose why. The goal is a more confident, more consistent instructor and instruction going to the AI, not a smarter AI.
See Hugo In ActionBook a demo and we will tailor the session to your institution's use case, whether that is large lecture courses, writing-intensive seminars, or STEM assessment.
