I built Signal because I kept hearing the same frustration from therapists: "I spend more time on notes than with my clients." But the deeper problem isn't just documentation — it's that therapists are flying blind during sessions. They're trained to notice micro-expressions, vocal shifts, emotional patterns... but they're also human. Things get missed, especially in high-intensity couples sessions. Signal sits alongside the session and does what a therapist's brain can't do simultaneously: track vocal biomarkers in real-time (pitch, energy, pause density, speech rate), detect safety risks across 6 clinical domains, and generate EHR-ready notes in one click. A few things I'm proud of: Not an LLM wrapper — we use ANFIS (neural-fuzzy) models for emotion detection, not GPT guessing from text Voice-first — we analyze HOW you say things, not just what you say Couples therapy — dyadic analysis with cycle detection (pursue-withdraw, criticize-defend, etc.) Usage-based pricing — 0.08/per processed voice minute No subscriptions. No minimums. HIPAA & PIPEDA aligned — built for real clinical use, not demos Would love your feedback. What would make you trust an AI tool in a clinical setting?

Signal by TheraPreneur

AI clinical assistant with real-time voice emotion analysis

AI clinical assistant with real-time voice emotion analysis