gramp — make sense of your health, privately
Your genome, health records, and the research that helps put them in context.
In active development. This page describes the current desktop work, including features newer than the public release.
A place for the whole picture
A DNA export, an Apple Health archive, a lab result, a paper worth reading: useful pieces of information that usually live in separate places. I'm building gramp to bring them into one encrypted workspace on your Mac, where you can explore what you have, follow new research, and ask questions with your own records as context.
Explore your health through anatomy
An illustrated body-system carousel connects the brain, heart, lungs, digestive system, liver, kidneys, and musculoskeletal system to relevant conditions and Apple Health trends. Swipe, scroll, or use the keyboard to move between systems and see the records linked to each one.
The view reflects what's on file. It distinguishes active conditions, items awaiting review, tracked measurements, and missing information. A system with no linked data isn't presented as healthy.
Import once. Keep the originals.
Drag in Apple Health or Health Auto Export files, or select several at once. Supported health uploads include ZIP, XML, JSON, and CSV. Gramp retains an encrypted copy of each accepted source inside its vault, so moving or deleting the original file doesn't break your history. Daily summaries and recent trends make the measurements easier to explore.
DNA imports support 23andMe, AncestryDNA, MyHeritage, and VCF files. Medical documents and notes add context for search and chat, while the Health Registry brings conditions and their supporting records together.
Research you can actually read
Research matches are enriched with available article abstracts and clinical-trial details before gramp writes a short, plain-language summary. It explains what the research found, what is still uncertain, and why a match might be relevant. Ongoing trials are described as research in progress, and research matches are kept distinct from diagnoses.
Sources include Europe PMC, which covers PubMed and preprints, and ClinicalTrials.gov. Manual summaries explain findings already on file; scheduled daily and weekly reports check for new research first. A local fallback can produce a simple summary without a configured AI provider.
Your vault, your choice of model
The database and retained files are encrypted on-device. Raw genotypes stay on your Mac. When using a cloud model, gramp prepares derived, de-identified context and records outbound payloads in a local audit. You can choose Anthropic, OpenAI, Google, or a local Ollama model.
Local inference works without a cloud AI provider; looking up new papers and clinical trials still requires an internet connection. Selective or full encrypted backups keep your records portable.
In progress: an iPhone companion
The companion work adds an encrypted bridge for viewing derived findings and reports on your phone, and sending approved Apple Health measurements back to the desktop. The first phone-to-desktop import path is implemented, but account setup, relay deployment, and a signed iOS build are still needed before it's ready for general use. Raw genotypes remain on the desktop.
Follow the project
Gramp is a personal project for organizing information and understanding research, not a medical device or a source of diagnoses. The public releases are earlier builds; the anatomy, retained uploads, and report improvements above are part of the current development work. For early access or a conversation about the engineering, get in touch.