Progetta
From course notes to lessons, from lessons to tests: an environment for designing teaching, planning lessons on the class calendar and assessing by competences.
Progetta (Italian for "design" or "plan") grows out of a conviction: designing your teaching, and assessing it by competences, is the most valuable part of a teacher's work and also the most demanding, so much so that it often stays on paper. Progetta uses artificial intelligence to remove the effort, not the thinking: the choices remain the teacher's, while the program does the maths, drafts the materials and keeps everything connected. Today it is my everyday working tool and it is not yet available to others.
From course notes to content
It starts from my course notes, in Word, LibreOffice or as web pages. Progetta reads them and breaks them down into units: explanations, definitions, examples, activities, exercises with solutions and typical mistakes. From then on, every piece of content can be reused in lessons, presentations and tests.
Lessons on the calendar
- Planning on the class's real calendar: days, periods, holidays, and even the position of the period in the school day, because attention in the fifth period is not what it is in the first.
- Lessons built in stages: explanation, exercise, activity, formative check, with timing kept under control.
- Presentations generated from the lesson, with links back to the course notes, video clips from one minute to another, and my interactive simulators attached to the slides.
- The week view: shows only this week's lessons, class by class, with the materials ready to open in the classroom.
Tests and assessment
- Tests that are designed, not improvised: a test blueprint, scores, balanced levels of difficulty.
- Four ways of judging the same test: by score, with the blueprint table, with an indicator grid, and by competences, so they can be compared and chosen knowingly.
- Competence-based assessment anchored to the Italian national curriculum guidelines (Indicazioni Nazionali) and to the European DigComp 3.0 framework: competences are broken down into observable evidence, from which four-level rubrics are built.
- Marking student by student, a class overview and printable individual reports, with the subject mark shown alongside the competence level.
A bank of exercises
Every exercise in the course notes becomes data: the evidence it assesses, its level in Bloom's taxonomy, its type, its difficulty. Difficulty is calibrated on the teacher's judgement: a few examples rated by me are enough for the AI to learn my scale. The typical exercise forms of the subject (in computer science: complete the code, find the bug, predict the output) are a repertoire the teacher can extend.
How it is built
A Python application that runs on my computer; the AI works only on request, and every operation shows its cost before it starts. Like all my projects: zero lines of code written by hand, but a long design process made of hundreds of small steps, each one tested before moving on to the next.