Challenge the Impossible.
A laboratory for real-world experiments at the intersection of AI, science, music and technology.
Different disciplines.
A common approach.
Build, test, measure, learn and iterate. Every project is designed around traceable inputs, explicit assumptions and reproducible results.
NeuroFly
AI-assisted behavioral and neural experimentation in Drosophila models, with a strict separation between simulation and biological evidence.
Explore project →MVA Rare Disease Lab
Reproducible variant prioritization and therapeutic hypothesis generation for Mosaic Variegated Aneuploidy.
Explore project →Music Lab
Interactive learning systems for chords, scales, rhythm, reading and instrument practice.
Explore project →Current Workbench
Tracking, synthetic state analysis and behavioral classification.
Variant interpretation, structural analysis and therapeutic hypothesis testing.
Interactive exercises, chord visualization and learning-system iteration.
Research is only useful when it can be explained and reproduced.
Define the question before running the experiment.
Track datasets, versions, parameters and source evidence.
Keep inputs and environments explicit enough to repeat the result.
Never mix observations, predictions, inference and clinical evidence.