Working documents and external material, organized so they can be added to continuously rather than rewritten.
Product requirements · user flows · UX specifications · feature definitions · product decisions.
Papers · research notes · expert conversations · literature reviews.
Architecture · API documentation · AI prompts · data structures · infrastructure.
Business model · pricing · market research · funding · pilot plan.
Supporter training · Supporter guidelines · safety protocols · escalation procedures.
Working documents are kept privately while the project is in early development. If something here is relevant to a conversation we're having, ask and I'll share it.
Each entry is recorded with a note on why it is relevant to InTune, rather than collected as a list of links.
Therapeutic transcripts, peer-support literature, training frameworks and professional standards.
Model capabilities, evaluation methods, and the safety literature relevant to sensitive contexts.
Work on assistance, augmentation and automation bias — central to keeping the Supporter in control.
Real-time communication infrastructure, speech-to-text accuracy and speaker identification.
Marketplace design, pricing, and digital health business models.
Session-transcript corpora such as the Stanford counseling and psychotherapy transcript collections, and the Alexander Street behavioral and mental health databases, are among the sources being reviewed for evaluating AI assistance against real therapeutic conversation.