ALFONSO  NAVARRO

Osteopath · Clinical AI Builder · Biomechanics

"Building AI systems that work in clinical reality —
not just in theory."

Osteopath · Physics · Clinical AI · Remote worldwide

Alfonso Navarro

About

Physics graduate (Universidad de Granada) and osteopath with 10+ years of clinical practice, including 2 years of intensive acute care during COVID-19 (2020–2022).

I build AI systems at the intersection of clinical reality and machine learning — not as a data scientist theorizing about health, but as a clinician who has worked under pressure and understands where systems fail.

Currently developing IO, a local-first autonomous AI agent for clinical autonomic monitoring, and working with Outlier/Anthropic on AI model evaluation.

Open to remote collaboration with clinicians, healthtech teams, and research groups working on wearable monitoring, autonomic dysfunction, post-infectious fatigue, or AI-assisted clinical decision support — regardless of location.

Projects

Published · Public repo · N=1 · Reproducible

ANS-Based Symptom Prediction in Post-Lyme Fatigue

N=1 Longitudinal Study · Consumer Wearable + AI Pipeline

48-hour prediction of symptomatic exacerbations using autonomic wearable data. ANS status predicts severe symptom days with 81% sensitivity (AUC=0.656). Sleep metrics show zero predictive value — the signal lives in the autonomic nervous system. Pipeline runs entirely local via IO. Scripts and methodology public.

Python · scikit-learn · LangGraph · Polar Grit X2 · ChromaDB

View on GitHub →
In progress · Local · Private · Clinical use case

IO · Local-First Clinical AI Agent

Autonomous Orchestration · RAG · ALMA Framework

Autonomous AI agent running locally on Mac Mini M4. Orchestrates LLMs (qwen3:14b + Claude) via LangGraph, enriches context with RAG/ChromaDB, and applies ALMA — an ethical-cognitive framework for clinical reasoning. Demonstrated use case: autonomic monitoring and 48-hour symptom prediction in post-infectious fatigue.

LangGraph · Ollama · qwen3:14b · ChromaDB · MCP · Python

See clinical use case →

Services

Clinical AI Consulting

Bridging clinical practice and AI implementation. Wearable data pipelines, autonomic monitoring systems, patient-reported outcomes integration. Built and validated in real clinical contexts. Available remotely worldwide.

Autonomic & Biomechanical Assessment

10+ years of clinical osteopathy and biomechanics. HRV-guided load management, post-infectious fatigue protocols, movement analysis. Evidence-based, measurable outcomes. Available remotely worldwide.

AI Model Evaluation (Healthcare)

LLM evaluation for clinical applications — safety, accuracy, and reasoning quality. Active evaluator for Anthropic/Outlier. Domain expertise in medical context and edge cases. Available remotely worldwide.

Contact

Interested in collaborating, consulting, or building something meaningful?
Reach out directly or connect on the platforms below.

Email: alfon.atman@gmail.com

LinkedIn: /navarro-kinetica-ai

GitHub: kinetica-IA