Looking for Collaborators

I'm donating my medical knowledge as open-source AI agents. Looking for experts who want to do the same.

I saw The Agency — 52k stars, 147 AI agents for business, all open-source. I talked to a colleague and we asked: why doesn't this exist for medicine? Doctors don't have the capacity. Computers can already do many of the tasks that burn them out. So I decided to start pooling what I know — and find others who want to do the same.

What I keep hearing

"Everything here is just bullshit now. Honestly. It's broken. This clinic for sure. But also the patients."

— ER physician, Germany. 2 AM. Just had his second child. Still on shift.

"I have 240,000 patients in my dataset. 20,000 are at risk for kidney failure. The doctors don't have capacity to change their behavior — even when the data is right in front of them."

— Health tech researcher. Real data, no system to act on it.

"A patient in Colombo gets a different standard of care than a patient in Boston — not because the knowledge doesn't exist, but because it's not accessible."

— Public health researcher, Johns Hopkins. Sees the gap every day.

"I spent 20 years training therapists. Now I could train agents that reach thousands — for free."

— Experienced psychotherapist and clinical supervisor, Germany.

What I believe

I've seen doctors burn out doing tasks a computer can already handle. I've talked to colleagues who have the data but not the capacity to act on it. I've worked with researchers in Sri Lanka and across Asia who have the expertise but not the tools.

I think if experts donate their knowledge as open-source agents, and patients can use them for free, something interesting happens: real-world data flows back. Research becomes possible that wasn't possible before. And the agents get better with every iteration.

The Knowledge Donation Loop

Experts donate knowledge → Agents are built → Patients use them for free → Users can donate anonymized data back → Researchers publish and improve the agents → Better agents → Better care → More data → Loop.

HOW I IMAGINE THE PATIENT LOOP

1
IntakePatient enters with symptoms, lab results, or screening data.
2
TriageRisk stratification. What's urgent, what's chronic, what's prevention?
3
DiagnosisSpecialized agents analyze labs, imaging, symptoms in parallel.
4
TreatmentEvidence-based protocol. Medication, lifestyle, follow-up — coordinated.
5
Prevention & CoachingWeight loss, nutrition, sleep, stress. Real programs, not generic advice.
6
MonitoringContinuous. Agents flag when something changes.
7
Patient walks out healthier. Loop repeats.

Departments

I'm thinking of it like a real clinic. Each department gets agents built by people who actually work in that field.

Prevention & Lifestyle Medicine Active

Weight Loss Coach, Nutrition Planner, Sleep Optimizer, Stress Manager, Exercise Programmer

Radiology & Interventional Active

Report Interpreter, Imaging Triage, DICOM Assistant, Intervention Planner

Mental Health Active

Trauma Therapy Guide, CBT Coach, ADHD Support, Burnout Assessor, Crisis Triage

Diagnostics

Lab Interpreter, Symptom Assessor, Screening, Differential Diagnosis

Treatment & Therapy

Treatment Planner, Medication Manager, Chronic Care Coordinator

Pharmacology

Drug Interaction Checker, Dosing Calculator, Side Effect Monitor

Research & Evidence

Literature Reviewer, Study Designer, Guideline Summarizer

Administration

Medical Coder, Documentation Assistant, Compliance Checker

What you get as a collaborator

Publish

I'm editor at APJLM (Asia-Pacific Journal of Lifestyle Medicine). We can propose special issues and fast-track relevant research.

Real data

We have colleagues with 240,000+ patient records. Not synthetic. Not simulated. Real clinical data for validation.

Network

Johns Hopkins, Shenzhen hardware ecosystem, Shanghai open-source community, Health Tech Meetups across Asia.

Own your domain

You become the lead for your department. Your name on the agents. Your expertise, open-sourced and cited.

China access

I'm based in Shenzhen with connections to Shanghai AI Lab, open-source associations, and government-backed health tech initiatives.

Early credit

The Agency went from Reddit thread to 52k stars. If this takes off, you're a founding contributor.

Early team

People I've talked to who share this frustration and want to build something about it. Currently recruiting across all departments.

Founder — Prevention & Lifestyle Medicine

Physician (MD). Based in Shenzhen. Editor at APJLM. Building OpenHealth. Organizing Health Tech Meetups across Asia. Medical AI hackathon winner.

Prevention Community China Bridge

Collaborator — Radiology & Interventional Radiology

Clinical radiologist. Deep-tech coder building DICOM plugins and multi-agent AI for medical imaging. Connected to major industry players and academic radiology AI networks.

Radiology DICOM/PACS Medical Imaging AI

Collaborator — Emergency Medicine & Care Pathways

ER physician and health tech researcher. Built autonomous treatment pathways for 240,000+ patients. Identified 20,000 at-risk patients through decentralized analytics. FHIR-native.

Emergency Medicine Care Pathways FHIR Real-World Data

Collaborator — Global Health & Lifestyle Medicine

Johns Hopkins Bloomberg School of Public Health. Associate editor at an international lifestyle medicine journal. President of an Asian lifestyle medicine council. Bringing world-class expertise to where it's needed most.

Global Health Lifestyle Medicine LMICs

Collaborator — Mental Health & Psychotherapy

Experienced trauma therapist and clinical supervisor. Trained therapists for 20 years. Now exploring how to train agents — scaling expertise that was previously limited to one-on-one supervision.

Psychotherapy Trauma Clinical Training

Looking for 5-8 more founding experts across all departments.

Why me

I'm not the expert in every department — far from it. But I've seen enough to believe this is worth trying. I'm a physician who codes, based in Shenzhen, with connections across Germany, China, and the open-source world. My role is to connect people, keep things moving, and learn from everyone involved.

I'm donating my own clinical and coaching knowledge first. Then I'm asking others to do the same. The system is free. The code is open. And if patients want to donate their anonymized data back for research — that's where it gets really interesting.

Want to build this with me?

Tell me what you know, what you'd build, and why this matters to you. No bureaucracy, no committee. Just people who want to make this work.

Reach out