Thesis analysis for academic work that checks evidence, not people
HiWi+ Thesis prepares the assessment: sources are checked for existence and currency, quotations matched against their location, and questions for the colloquium suggested. Assessment and mark remain entirely with the examiners.
HiWi+ Thesis is developed and is in practical testing with real theses. We say so openly, because trust and academic integrity come first in a final examination. Anyone who wants to be involved early can join as a pilot partner and test the module in practice.
Where thesis analysis relieves the assessment work
Not in the academic judgement, that remains the examiners' task. But in the groundwork before it: looking up sources, checking quotations, comparing formalities.
Check that sources exist
Every entry in the bibliography is matched against catalogues, DOI registries and publisher records. Invented titles and DOIs that do not resolve stand out before anyone looks them up.
Match quotation and location
Where the source is accessible, it is checked whether the quoted statement actually appears at the given place. That is the check for which there is rarely time in a report.
Currency of the literature
It becomes visible how old the body of work drawn on is and whether newer work exists on central points. A finding for the conversation, not an assessment.
Completeness of evidence
Supported and unsupported statements are set side by side, as are bibliography entries that appear nowhere in the text.
Questions for the colloquium
For the load-bearing sections, suggested questions are produced with reference to the specific passage. Which one is asked is decided by the examiners.
Follow the methodology
The described approach, the data used and the conclusions drawn are laid side by side, so gaps in the argument can be addressed in the conversation.
Check formal requirements
Citation style, length, required components and declarations are checked against the rules of your institution. Formalities should not hold up the report.
Prepare the second opinion
Both assessors receive the same reference overview. The conversation then starts with the academic questions instead of reconstructing the citation apparatus.
How a reference overview comes about
Four steps, and the fourth always belongs to the examiners.
Store the requirements
Examination regulations, citation guidelines and the requirements of your department are set up once. The module checks exclusively against that specification.
Read in the thesis
The submitted version is read in and broken down into its parts: text, bibliography, appendices and declarations.
Match the evidence
References are checked against catalogues and registries and, where possible, matched with the quoted location. Every finding names the passage it refers to.
Examiners decide
You receive a reference overview and suggested questions. What goes into the report or the colloquium is determined exclusively by you.
The principles we do not depart from
In a final examination a tool can do particular damage. That is why these points are anchored in how it is built, not stored as a setting.
What belongs to the module
Reference overview
All references with their check result: resolvable, not found or quoted differently, each with the passage concerned.
Suggested questions
Questions on the load-bearing sections, referenced to the passage in the thesis. Marked as a draft and freely editable.
Formal check
Comparison against the citation guidelines, length and required components from your examination regulations.
Audit log
Which checks ran, with time and version. That holds up when an assessment has to be justified later.
What HiWi+ Thesis deliberately does not do
A tool that promises more than it can be held to in a final examination does more harm than good.
Karlsruhe University of Applied Sciences
HiWi+ is in use at Karlsruhe University of Applied Sciences and was developed together with the departments there, along real workflows rather than at a desk.
“Our main goal was to reduce administrative processes and workload within the faculty while at the same time offering students a better service, for example with the faculty chatbot they can ask questions about their studies around the clock. The HiWi+ modules were developed for our specific needs at the faculty, but they are certainly just as usable for many other educational institutions.”
What institutions ask before rolling it out
No, and that is a deliberate decision. Tools that output a percentage for AI use are demonstrably unreliable and particularly disadvantage students who are not writing in their first language. HiWi+ Thesis instead checks what is objectively verifiable: whether the sources given exist, whether they can be found and whether the quotation appears at the stated place. Invented sources are the more robust finding anyway.
No. It produces neither a mark nor a grade recommendation nor an assessment of independent authorship or subject understanding. It checks statements in the document and suggests questions for the colloquium. Assessment and mark remain entirely with the examiners.
Assessing learning outcomes and detecting prohibited behaviour during examinations are named as high-risk areas in the EU AI Act. That is precisely why HiWi+ Thesis is scoped so that it does neither: it produces no statement about a person, it checks evidence in the document and suggests questions. The intended purpose with this exclusion list is part of the operating instructions.
The examiners. The suggestions are marked as drafts and freely editable. They are preparation material, not examination rules.
They are processed at runtime and are not fed into a model. There is no learning loop on student texts. On request, processing runs in your own data centre or entirely locally via HiWi+ Local AI. Retention periods are agreed with your department.
HiWi+ Thesis is developed and is in practical testing with real theses. We say so openly, because trust and academic integrity come first in a final examination. Anyone who wants to be involved early can join as a pilot partner and test the module in practice.
Yes, and that is the sensible route. One department, one citation guideline, one examination period. After that you can judge whether the reference overview genuinely relieves the assessment work.
Test it on one of your own theses
We demonstrate the module on a thesis from your department that has already been assessed. You then see on a real case how robust the evidence check is, before anything is decided.