thetahealth/mirobody

The AI-native health data engine — collect, standardize, and reason over labs, wearables & genomics.

What it solves

Mirobody is an AI-native health data engine designed to solve the problem of fragmented and inconsistent health data. It translates diverse health readings—from lab reports, wearables, and genomics—into a single, standardized language (LOINC, UCUM, and FHIR) that AI agents can accurately read and reason over, regardless of the original language or unit of measurement.

How it works

The system operates in three primary stages:

  1. Collect: Pulls signals from various sources, including device providers, SQL sources, multiple file formats, and Apple Health.
  2. Translate: Standardizes readings by resolving terms in any language to canonical codes (LOINC, SNOMED CT, RxNorm) and normalizing units to UCUM.
  3. Answer: Employs an AI agent that reads original documents through a virtual filesystem to provide answers, charts, and citations based on the standardized data.

Who it’s for

It is intended for developers building health-AI applications, caregivers, and organizations needing a self-hosted, standards-based infrastructure for health data management and AI reasoning.

Highlights

  • Multilingual Resolution: Resolves health terms in English, Chinese, Japanese, and several other languages to a single LOINC code offline.
  • Standards-Based: Produces FHIR-ready records using LOINC 2.82, SNOMED CT, and RxNorm.
  • AI-Agent Integration: Includes an agent harness that can be served via MCP to tools like Claude Desktop and Cursor.
  • Self-Hosted: Can be deployed locally using Docker with a full stack including Postgres, pgvector, and Redis.
  • High Coverage: Covers 211/211 panels found in ordinary checkups across multiple languages.

Related

  • Project
    Mengqi97/chinese-medical-dataset

    A curated collection of open-source Chinese medical datasets for tasks like medical QA, entity recognition, and knowledge graph construction to support the development of medical AI.

  • Project
    ZeroTang05/cyber-doctor

    A multimodal AI health assistant that provides disease diagnosis and medical record analysis using RAG, knowledge graphs, and voice interaction.

  • Project
    the-momentum/open-wearables

    An open-source platform that unifies wearable device data from multiple providers into a single API, enabling AI-powered health insights and automations.

  • Project
    aipoch/medical-research-skills

    A curated library of 550+ medical research agent skills for AI agents to assist biomedical researchers with evidence discovery, protocol design, data analysis, and academic writing.

  • Dispatch
    Google DeepMind AI Co-Clinician Research Initiative

    Google DeepMind has introduced the AI co-clinician research initiative, a multimodal AI system designed to support healthcare providers through evidence synthesis and real-time patient interaction under clinical supervision.