Beyond the scribe
On the surface, Sigmund is a scribe — he writes the note, captures the MSE, cites the evidence. But the structured understanding he builds of each patient is the foundation for something larger: helping clinicians deliver demonstrably better care, condition by condition, and carrying that care into the weeks between visits through the PAWS wearable. The scribe makes the work easy. What it makes possible is the point.
ADHD
The evidence for ADHD is unusually clear and unusually often ignored. A child with comorbid anxiety should not start where an uncomplicated case starts. An adolescent two years into a stimulant trial is a different decision than a treatment-naive eight-year-old. The right answer depends on population, comorbidity, prior response, and sequence — and a clinician working in fifteen-minute blocks cannot reconstruct the relevant trial for every profile.
Sigmund weighs the specific profile against the actual literature — his validated CEBA engine ranking treatment — behavioral parent training, non-stimulant, stimulant, in the order the guidelines support — and he surfaces the citation chain so the clinician can see why, override freely, and document a defensible choice. The result is not faster notes. It is care that matches what the evidence actually says, for this patient, in the moment the decision is made. See the clinical decisions in detail →
Depression
Medication is half the story in depression. Behavioral activation — re-engaging the patient with rewarding, values-aligned activity — is among the most robust behavioral treatments there is, and it fails for one mundane reason: it lives entirely between visits, where the clinician cannot see it. A plan made in clinic on Tuesday is invisible by Thursday.
Sigmund carries behavioral activation home through the PAWS wearable, staying present between visits. The activity schedule built with the clinician travels home with the patient; PAWS prompts the planned activity, captures whether it happened, and tracks activation and mood as they move. The clinician titrates against real behavior instead of recall — sees which activities lifted mood and which did not, and adjusts the plan with data, not a self-report colored by the depression itself. The behavioral half of treatment becomes as measurable as the pharmacologic half.
Anxiety
The frontline treatment for most anxiety is not a benzodiazepine — it is cognitive-behavioral skill and a regular mindfulness practice. Both are skills that compound with practice and decay without it, and both happen at home, between sessions, where they have historically been impossible to support or measure.
Sigmund extends mindfulness-based practice and CBT skills — cognitive restructuring, graded exposure, paced breathing — through the companion, and uses PAWS to track engagement and physiological response: the practice actually done, the arousal that fell with it. The clinician arrives at the next session knowing what was practiced and what worked, and the patient builds the durable skills that carry anxiety treatment far beyond the prescription.
Schizophrenia
Schizophrenia is the hardest case for continuity, and the clearest test of whether software helps care or just documents it. The most effective treatment — clozapine, the only antipsychotic shown to reduce suicidality — is also the most underused, defeated by its monitoring burden. And relapse builds silently between visits.
Sigmund holds the clozapine documentation tail automatically, so the medication that works stays in place, and watches through PAWS to surface the early-relapse signal — sleep, activity, social rhythm — while there is still time to act. The outcome that matters is not a symptom score; it is a life kept intact. Read the full schizophrenia model →
The PAWS connection
Every one of these gains shares a mechanism: care does not stop when the visit ends, and neither does Sigmund. That continuity runs on PAWS — a working wrist-worn companion that pairs passive sensing with an LLM-driven therapeutic agent on the Google Pixel Watch. Built in collaboration with Google and advancing through an NIDA UG3/UH3 award toward clearance as a software-based medical device, PAWS has already outperformed a general-purpose chatbot on clinical fidelity in prototype testing and delivered the correct safety response in every high-risk scenario it was tested against.
Through PAWS, Sigmund stays present between visits — reading the ecological signal continuously and in context — sleep, activity, circadian and behavioral rhythm, physiological arousal, and the adherence the patient could never accurately recall in clinic. The companion platform that proved itself in young-adult substance-use care is the same one that now carries behavioral activation, CBT practice, and relapse monitoring into the weeks between visits.
That continuous signal is what turns a fifteen-minute monthly snapshot into a longitudinal trajectory. The clinician sees the depression lifting before the patient can name it, the relapse building before it lands in an emergency room, the exposure practice that actually lowered arousal. The loop is simple and complete: Sigmund captures the encounter, PAWS monitors the life between encounters, the signal surfaces at the next decision, and the clinician adjusts.
This is the whole thesis. Sigmund makes documentation effortless for the clinician — and uses what he understands to help them deliver care that is measurably, continuously better for the patient.
Sigmund is investigational and is intended to assist — not replace — clinical judgment. The clinician directs every treatment decision; the patient signs nothing away. Built by the Sultan Lab for Mental Health Informatics at Columbia University Irving Medical Center.
