‹ Back to projects

Repetita

Study material that adapts to you, not the other way around.

Repetita is a virtual assistant that preps students for oral and written exams by simulating the real thing: you feed it material — text, a photo, a PDF, or just the topic's name — and it runs an exposition with questions of rising difficulty, voice input and live feedback while you talk.

Material can also come from an image or a multi-page document: multi-model OCR transcribes and cleans it up into notes and summaries ready for review. Every session ends with a score from 1 to 10, plus what went well and what to revisit.

Behind the scenes it uses several AI models together rather than one fixed choice, picking whichever fits the step at hand — reading an image, judging a spoken answer, writing a summary.

  • Platform
    Web app (PWA) + native Android app
  • AI
    Gemini, GPT-4o, DeepSeek V3
  • Simulation
    Oral exam with real-time feedback
  • Languages
    Italian, English, French, Spanish, German, Ancient Greek, Latin
  • Simulates an oral or written exam, with questions of variable difficulty
  • Reads text from photos and PDFs with multi-model OCR, including handwritten notes
  • Generates notes, summaries and study material from text, images, or just the topic's name
  • Listens to and grades a spoken answer through voice recognition
  • Returns a score from 1 to 10 with strengths and topics to revisit
  • Includes a marketplace for used school textbooks, searchable by scanning the ISBN