Medicine & Health
Health information rarely exists as a single observation. Symptoms change. Treatments evolve. Different professionals contribute different perspectives. Laboratory results acquire meaning when compared over time. Unexpectech explores software, structured information, semantic systems, and artificial intelligence that preserve those connections instead of flattening them into isolated records.
Core principle: Technology should increase understanding without pretending to replace clinical expertise, veterinary judgment, patients, caregivers, or the people responsible for care.
Current work
The Medicine & Health work spans a functional prototype, veterinary system design, health-data research, and a cross-area person-centered AI project. Each page separates what has already been implemented from what is still being explored.
Functional Medicine
A software experiment that turns the Functional Medicine Matrix into structured, editable information and explores how it could become longitudinal rather than remain a static clinical form.
Explore SmartMatrix →Veterinary Medicine
Research into continuity of veterinary information: consultations, owner observations, clinical notes, hospitalization follow-up, veterinary knowledge retrieval, and contextual support across an animal's history.
Explore Animal Care →Information Architecture
Exploration of interoperability, provenance, semantic retrieval, longitudinal records, and the rules needed when AI operates over medical or veterinary information.
Explore Health Data →rAIn contributes a person-centered health dimension: consultations, medications, doctors, therapy, documents, reminders, event capture, transcription, structured summaries, contextual retrieval, and controlled sharing with trusted collaborators.
Research principles
Across these projects the same engineering problem keeps appearing: information has a source, a time, a reason for existing, and relationships with what came before and after.
Preserve how symptoms, observations, treatments, results, and decisions change across time.
Keep the source, author, date, event, and transformation history of important information visible.
Connect information to the people, events, documents, and previous history needed to understand it.
Permissions, consent, sensitivity, and appropriate access remain part of the system—not an afterthought.
AI should first recover the available evidence and should distinguish retrieved facts from generated summaries or inferences.
Human and veterinary medicine share information problems, but their workflows, terminology, standards, and clinical interpretation cannot simply be treated as identical.
The research question
A consultation can remain connected to its questions, audio, transcript, summary, recommendations, medications, documents, and pending follow-up. A laboratory result can be understood beside earlier results. An observation can preserve who recorded it and why it mattered. A veterinary note can remain connected to owner observations and later outcomes.
Context before automation.
Traceability before prediction.
Understanding before generation.
Canonical Project Index
Cross-area in Medicine & Health
