~/projects/arianna
Arianna
Arianna takes a course's scattered slides and question sets and rebuilds them into a single, well-ordered manual — with exam questions woven in section by section. Read it on the web, author it from a macOS app.
The problem
Arianna started as a tool for myself. When I began studying psychology, the course material was what it usually is at university: slides in PDF or PowerPoint, handed out by the professors running the course — and often badly written, disorganized, incomplete, sometimes wrong or simply out of date.
On top of that, courses tend to ship two disconnected things: a set of study material that may be incomplete, and a separate set of example questions that may or may not line up with the lessons. Nothing ties them together, and there's no clear picture of whether you've actually covered what the exam will ask.
So instead of studying from the raw slides, Arianna rebuilds them: it takes everything the course provides and produces a single, properly structured manual to study from.
My role
I'm the sole founder of Arianna — I designed and built it end to end, across the macOS authoring app, the web reader and the mobile reader.
It began as something I needed myself: a better way to study once I'd started a psychology degree and hit the limits of the material the courses handed out.
How it's built
Arianna doesn't wrap AI features around a generic reader — it rebuilds the study material itself, then serves it across three clients.
From scattered sources to one manual
Arianna ingests all of a course's exam materials and builds a brand-new manual from them — well-ordered and didactically structured, so topics are presented in the right sequence and at the right level of depth.
Questions and quizzes woven in
While the manual's macro-structure is built, each important section gets its matching questions attached. Courses usually provide an incomplete study set and a separate, loosely related question set; Arianna interlaces both. On top of that, the app runs randomized quizzes — drawn from the source questions and generated on the fly by its own AI.
A manual-grounded AI assistant
An AI assistant tied closely to the manual gives clarifications on the study material, with direct citations that point back to the specific sections they come from — so an answer always traces to where the manual actually says it.
Progress and competence tracking
Arianna tracks which sections you've read and lets you flag what to review later. A competence score per chapter, built from your correct and incorrect quiz answers, shows where you're strong and which chapters need another pass.
Three clients, one system
A macOS desktop app — built in Rust, with parts in Python — authors the manuals. A web reader, built on Ruby on Rails with Inertia, is live. A React Native mobile reader is in development. The macOS builder is the app I use myself to structure the didactic process.
AI-built manuals from any topic
With the macOS app you can also build a study manual on a topic of your choice, with no certified source material at all — a more abstract mode where the AI does most of the work.
Flashcards (planned)
A future feature: generating flashcards from the key points identified in each chapter or paragraph, to be consumed in both the web and mobile readers. It isn't built yet.
By the numbers
24–72+
lessons per reference course
150–1,000
exam questions per set
3
apps: macOS builder, web + mobile readers
Screenshots
AI assistant answering with a citation into the manual
[ASSET NEEDED: screenshot of the manual-grounded assistant showing an answer with a direct citation to a specific section]
Per-chapter competence overview
[ASSET NEEDED: screenshot of the competence tracking showing strong and weak chapters]
Authoring a manual in the macOS app
[ASSET NEEDED: screenshot of the macOS builder assembling a structured manual]
Status
The web reader is live at arianna.marzapower.com. The React Native mobile reader is in development, and the macOS app — the tool I use to author the manuals — drives the content pipeline. Flashcards are the next feature planned.
Related writing
Technical write-ups on Arianna are on the way.