How DigiBot 2.0 works
A short tour of the artefact, what it measures, what the AI does, and what your data is used for.
- Fill in your profile so we can personalise the assessment to your role and tasks.
- Take a quick (10 items) or in-depth (34 items) AI-literacy assessment based on the MAILS framework.
- Review your results - an overall percentage, three category scores, and nine sub-dimension scores.
- See ranked recommendations: where to invest first, with concrete next-morning steps and supporting evidence.
- Optionally build a training plan; we generate short, personalised exercises for each session.
The assessment is based on the MAILS scale (Carolus et al. 2023). MAILS covers three categories - Use & Apply AI, Understand AI, and Detect AI - across nine sub-dimensions and 34 validated items.
Each item is a Likert statement (1 strongly disagree to 5 strongly agree). Your raw responses are stored alongside your account; only aggregated scores are shown anywhere.
An LLM rephrases each item so it sounds natural for your role and tasks; the underlying validated wording is preserved.
A retrieval-augmented system (RAG) pulls evidence chunks from a curated AI-literacy knowledge base and ranks them per dimension; each chunk shows a confidence band.
Training exercises are generated from canonical templates, then localised and personalised by the LLM under per-type integrity constraints.
Profile, responses, recommendations and training progress are stored in a Cloud SQL database hosted in Switzerland (europe-west6). Access is restricted to the artefact backend.
We do not sell, share, or train external models on your data. The thesis evaluation uses pseudonymised aggregates only.
For questions, account issues, or to be removed from the allow-list, contact digibot.bfh@gmail.com.