All Work

Tiny Shifts

Clinical psychologist, AI co-founder, HIPAA-adjacent infrastructure — coaching memory that persists where human coaches can't.

Tiny Shifts platform

Context

Elisha Goldstein is a clinical psychologist with 20+ years of practice. The problem: clients leave sessions with insight and lose it by Wednesday. Coaching works in weekly 50-minute windows; behavior change requires constant repetition.

Strategy

A coaching reinforcement platform that converts session recordings into persistent behavioral memory, then delivers timed nudges between sessions based on what each client is actually working on.

Solution

Design partner phase active with two top-tier US coaching organizations. Enterprise tier at 85% margins.

Technical Architecture

Recording layer (Recall.ai). Extraction layer (OpenAI against structured schema co-designed with Goldstein). Memory layer (FastAPI + SQLAlchemy + PostgreSQL, cross-session pattern detection). Nudge delivery timed to commitment structures from each session.

HIPAA Considerations

Platform operates in HIPAA-adjacent territory. Architecture minimizes PHI surface: transcripts processed into behavioral abstractions. Does not claim HIPAA compliance — designed to support compliant deployment configuration.

Business Model

Top 100 US coaching organizations. $85/user/month enterprise. 85% margins. Design partners co-develop feature priorities for preferential pricing.

20+
Sessions tracked per client
85%
Margins at enterprise
2
Design partners active
Cross-Session Memory B2B SaaS Behavioral Nudges Enterprise FastAPI PostgreSQL Recall.ai