One SLP, thirty patients.

A'ber is a digital speech therapy platform for speech language pathologists treating dysarthria. The adaptive clinical engine handles session to session exercise selection, intensity calibration, and per dimension severity tracking, freeing the SLP to focus on clinical judgment and intervention planning instead of the mechanical work of building the next session by hand. A single clinician can carry a caseload of thirty or more dysarthria patients on A'ber while preserving the per patient quality a typical eight patient caseload provides.

Longitudinal severity tracking

Every recording feeds into a per dimension severity trajectory. The SLP sees how loudness, articulation, prosody, and the other sub tracks move week over week, with confidence intervals derived from a Kalman filter rather than a single noisy point estimate. Trajectories surface plateaus and regressions early so intervention happens before the patient notices the slide.

Flagged events, not noise

A'ber does not page the clinician on every session. Flagged events fire when the system detects a meaningful deviation from the patient's own baseline: two consecutive sessions missed, a sustained drop in clarity beyond the calibrated noise floor, or a confidence collapse on a previously stable dimension. The SLP receives a structured prompt with the relevant data attached, not a generic alert at two in the morning.

The caseload view

The caseload page is the SLP's home. Each patient row shows severity at a glance, current focus sub tracks, last session recency, session count, average clarity, a sparkline of the clarity trend, and any active alerts. Rows needing attention are tinted and ordered above the rest, so triage happens in seconds rather than minutes.