Music education, made for Uganda.
Music-Forte is two systems in one workspace: a chat-driven AI assistant aligned to the NCDC 2025 syllabus and UNEB Practical Music, and FORTE Notation โ a tri-modal manuscript editor where score, ABC source, story, and phrase commands edit the same piece.
Why we built this
Most music software is built for conservatoires and recording studios in New York or Berlin. It assumes a desktop, a piano, a steady connection, and a teacher trained at a private music college. That isn't the reality for an S5 music teacher in Mbarara or a candidate preparing for UNEB Practical Music in Gulu.
Music-Forte starts from the other end. Tools that work on a phone, a tablet, a shared school computer. Lessons that map to the actual topics on the NCDC syllabus. Set works and traditions that include baakisimba, bwola, embaire, and a Mukasa hymn arrangement โ not as a token but as set works the system can compose, harmonise, and analyse on the same footing as Bach or Mozart.
Two systems, one workspace
Music-Forte AI
The AI surface is a chat. Four modes โ Guest, Teacher, Student, Expert โ change which tools are available and how the assistant interprets ambiguous requests. Teacher mode reaches into the NCDC corpus and emits lesson plans, schemes of work, scenarios, and rubric-anchored feedback. Student mode is drill-shaped, with intervals, cadences, sing-back, and dictation. Expert mode opens the music-generation models โ choral baroque, hymn SATB, folk monophonic โ and lets a composer iterate on variants.
FORTE Notation
FORTE Notation is a manuscript editor with four ways of editing the same piece, each fully reversible:
- Score โ direct manipulation on the engraved staff.
- Code โ ABC source as plain text with a notation-aware keypad on mobile.
- Story โ natural language. "Make this sadder." The model proposes a diff; you accept, edit, or reject.
- Phrase โ deterministic command grammar. "At bar 5 beat 1, insert quarter D." Parser-driven, no model, fully auditable.
Other surfaces fold in around the four modes: a Voice mixer, an Inspector that flags voice-leading issues, comments anchored to notes, live multi-user collaboration, hum capture and contour sketching for melody intake, and Practice mode โ where the score listens to you play and follows along.
Who Music-Forte is for
- Music teachers in Ugandan secondary schools โ especially those preparing students for UNEB Practical Music or working through the NCDC 2025 syllabus.
- Students at S5โS6 and earlier, drilling intervals, sight-reading, ear training, and analysing set works.
- Composers and arrangers who want a manuscript editor that handles equal-temperament and the equidistant tunings of Ugandan traditional instruments (amadinda, embaire, adungu).
- Choirs and arrangers of liturgical music โ Mukasa-style hymn settings are first-class in the corpus, not an afterthought.
Music-Forte runs on phones, tablets, and desktops. The mobile shell is purpose-built โ not a shrunk desktop โ and the tablet picks landscape or portrait based on orientation. Schools share devices; the design assumes that.
Principles
- Reversibility before generation. Every AI proposal in FORTE is a diff you can accept or reject. We never overwrite without permission.
- Determinism where possible. Phrase mode + the command-editor parse 78+ tested edit patterns without ever calling a model. Faster, cheaper, auditable.
- On-device when reasonable. Pitch detection, polyphonic following, and the practice audit run in the browser. Your audio doesn't leave the device unless you opt to share.
- Honest empty states. No fake content. If you have no lessons, the library says so. If your streak is zero, we say zero.
- Curriculum on equal footing. Western tradition and Ugandan tradition trained side-by-side. The same composition tools handle a fugue and a baakisimba.
Contact
Reach us at support@music-forte.app. For privacy, security, or data-rights requests, see the privacy policy.