LalaScript is a browser-based music notation platform that lets musicians upload a PDF score or create one from a prompt, then edit it using plain-English instructions. The engine performs semantic music analysis, preserving melody while applying constraint-based transformations like transposition, simplification, and re-voicing for playability. Users can hear results, download files in MusicXML, PDF, MP3, or MSCZ formats, and even hum a tune into their mic to generate editable notation.
Funding
Funding not disclosed
Founders
Product
Problem
Musicians frequently need to adapt existing scores—transposing for a singer's range, simplifying parts for a student's skill level, or rearranging instrumentation—but traditional notation software requires steep learning curves, time-consuming manual editing, and desktop installation. This creates friction for teachers, accompanists, students, and performers who need quick, accurate adjustments.
Solution
LalaScript is a browser-based platform that converts uploaded PDF scores or generated pieces into structured musical data, enabling users to edit notation using plain-English commands. The system tags every note with its musical role—melody, inner voice, or bass—and applies constraint-based transformations that treat hand span, chord density, and rhythmic complexity as hard limits. Users can request changes like "simplify for a beginner" or "transpose to D major," and the deterministic rule-based engine executes exact, repeatable edits while reserving AI only for creative tasks like melody generation. Scores can be played back, further edited, and exported in MusicXML, MusicXML-compatible, PDF, MP3, or MSCZ formats, with generation features including prompt-based composition and microphone melody notation.
Target Audience
Primary users are music teachers, accompanists, musical directors, singers, and students who need to quickly adapt, transpose, simplify, or rearrange existing scores without mastering complex desktop notation software.
Features
- Upload PDF scores and edit them via natural language text commands referencing specific bars, staves, beats, or note positions
- Semantic analysis engine that tags every note by musical role (melody, inner voice, bass) before transformation
- Constraint-based output engine treating hand span, chord density, and rhythmic complexity as hard bounds for playable arrangements
- Deterministic, rule-based operations for transpose, re-voice, and chord thinning; AI used only for creative edits
- Prompt-based composition and microphone input that turns hummed melodies into editable, theory-correct notation
- Export options to MusicXML, PDF, MP3, and MSCZ, alongside built-in audio playback and sharing capabilities