Brainif is the operating system for closed-loop biological neural experiments — design, simulate, replay, and analyze living neural systems.
Pick an input pattern and run the stimulation — watch the signal travel encoder → electrodes → synthetic activity → decoder → feedback.
Everything from the experiment spec to the decoded result lives in one governed, reproducible pipeline — the same whether the activity is simulated or replayed.
Compose hypotheses, tasks, encodings and controls in a versioned, reproducible spec.
Assign input, output and feedback roles across a multi-electrode array.
Route decoded activity back into stimulation — open, closed, reinforcement or homeostatic.
Import recordings, map fields, and replay them through the same analysis pipeline.
Per-channel rates, burst detection, latency and stability from any run.
Feedback vs. no-feedback, baseline vs. adaptation — side by side, descriptively.
Every result traces to its config version, simulator, decoder, seed and file hash.
A vendor-neutral interface ready for live MEA and remote biological-computing backends.
Every run is labeled with the mode that produced it — so a result is never mistaken for something it isn't.
Test experiment logic using synthetic neural activity.
Import and analyze recorded biological data.
Connect to supported laboratory or remote biological-computing systems.
Electrode map, live raster, decoder confidence and run metadata — all in one governed view.
Brainif is a software platform and research control plane. It does not claim that simulated neurons are living neurons, and it never infers consciousness.
Every experiment is a versioned object. Define it in the studio or POST it directly — the same governed pipeline runs either way.
POST /api/experiments
{
"name": "Binary signal classification",
"mode": "simulation",
"inputEncoding": "spatial_stimulation",
"decoder": "firing_rate",
"feedbackProtocol": "closed_loop"
}Replay a dataset, design a protocol, or talk to us about connecting live hardware.