Project Info
This project did not submit a demo video on Devpost.
Inspiration
The inspiration came from trying to think of ways to have a real world impact when we started thinking about the fact that we didn't actually know how to easily access our medical records. One team member has to go through four portals himself just to figure out which record is the correct one.
What it does
MedMeld allows patients to pull their medical history from their respective providers in one place. It then makes understanding your history easier by removing the medical jargon with the help of Google Gemini
How we built it
We built this project with python using fetch.ai's uagents to perform tasks like pulling the data from the respective health care providers. We use gemini API to summarize and aggregate medical records from various sources for easy viewing and understanding of complex medical jargon
Challenges we ran into
We had trouble figuring out how to get the uagents to communicate with each other over the net. We also had trouble connecting to our postgres database
What we learned
We learned about distributed communication with fetch.ai and integrating with google's API
What's next
We plan to significantly enhance the user experience by making the process of accessing and managing medical records more intuitive and user-friendly. Our vision is to empower patients to take control of their healthcare data with ease and confidence. By integrating Google Gemini, we can add intelligent features that provide patients with more comprehensive insights into their medical history, while maintaining the highest standards of privacy and security.
This is the CalHacks 11.0 project of:
Sulaiman Mulla: sulaiman_1@tamu.edu Jeffrey Cheung: jcheung@tamu.edu Alex Beamer: alexbeamer@tamu.edu Gabriel Gonzalez: gabriel29@tamu.edu
Analysis
View
Metric
- 22
- 11
- 10
- 9
Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.
Technology
- PythonIn code
- SQLIn code
- Google GeminiClaimed
2 of 3 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.
AI coding agents
No AI coding agent signals were found in this repository.
Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.
Codebase size
Source size
52 KB
Source files
16
Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.
Repository
CheJeff/MedMeld
23 files · 57 KB · @ d65553e
Structure
Application logic
11 files · 48%Domain rules, services and shared utilities.
Supporting
Layers are inferred from where files sit in the tree, not from reading the code. A project that names its directories unconventionally will read oddly here — open the file browser to check anything the diagram implies.
Languages
- Python97%
- SQL3%
- Markdown1%
Share of indexed source by file size. Binary and vendored files are excluded.
Dependencies
requirements.txt
pypi · 6- faker
- google-generativeai
- psycopg[binary,pool]
- python-dotenv
- reflex
- uagents
gui/requirements.txt
pypi · 1- reflex
messages/pyproject.toml
pypi · 1- uagents
Declared in the repository’s manifests at the indexed commit. A declared package is not proof it is used, and runtime dependencies are listed first.
Feature verification
Distributed agent communication via fetch.ai uAgents to pull data from providersVerified
We built this project with python using fetch.ai's uagents to perform tasks like pulling the data from the respective health care providers
Claimed on Devposthigh confidenceagents/hospital.py:8— Defines a uAgents Agent with its own port/endpoint that listens for PatientQuery messages and replies with PatientDataagents/identity.py:11— Separate uAgents identity agent handling account creation, sign-in, and provider linking messagesagents/test/identity_agent.py:6— A test uAgent that sends ReqSignIn/ReqAddProvider messages to the identity agent's address, demonstrating actual agent-to-agent messaging
Dummy/mock hospital data generation for demoing multi-provider pullVerified
pull the data from the respective health care providers
Claimed on Devposthigh confidencedummy_data/hospital1Generator.py:8— Script populates a hospital1_records.db sqlite database with hardcoded fake patient records used by agents/hospital.py
PostgreSQL-backed patient/account data storageVerified
We also had trouble connecting to our postgres database
Claimed on Devposthigh confidencescripts/dbsetup.sql:5— SQL schema defining Patients, Providers, and PatientProviders tables for Postgresagents/identity.py:36— identity agent uses psycopg.connect against PGSQL_CONSTR env var to insert/query the Patients table
Reflex-based web GUI for login, profile, and Q&A on medical recordsVerified
Built with reflex
Claimed on Devposthigh confidencegui/MedMesh/MedMesh.py:9— Imports reflex, defines rx.State and multiple rx.App pages (login, profile, record, create-account, healthcare-providers, forgot-password)
Pull medical records from multiple healthcare providers into one placeCode-supported
MedMeld allows patients to pull their medical history from their respective providers in one place
Claimed on Devpostmedium confidenceagents/hospital.py:59— Hospital1 uagent handles PatientQuery and returns PatientData read from a sqlite databaseagents/hospital2.py:54— Hospital2 uagent handles PatientQuery and returns PatientData from a separate sqlite database, showing two distinct provider sourcesmessages/__init__.py:62— Shared PatientQuery/PatientData message schema used by both hospital agents for uagents communication
Simplify medical jargon using Google GeminiCode-supported
It then makes understanding your history easier by removing the medical jargon with the help of Google Gemini
Claimed on Devpostmedium confidencegemini.py:9— generate_info() calls Gemini's gemini-1.5-flash model with a prompt asking it to summarize the medical record before answering, matching the jargon-simplification claimgui/MedMesh/MedMesh.py:68— generate_answer() calls gemini.generate_info but json_input is hardcoded to an empty string, with the real data call commented out (bureau.get_combined_patient_data), so no actual patient record is ever passed to Gemini
User login authenticationCode-supported
Patients access their account to view medical history (implied by login/profile pages and identity agent sign-in)
Claimed on readmemedium confidencegui/MedMesh/MedMesh.py:20— The GUI's login() method checks against hardcoded credentials ('Jane222'/'ILoveToHack') rather than calling the identity agent's ReqSignIn flow, so the real auth backend exists in agents/identity.py but is not wired into the frontendagents/identity.py:49— A working sign_in handler exists that verifies hashed passwords against Postgres, but it is only exercised by the separate test agent, not the GUI
Aggregation of pulled records into a single unified viewClaimed only
MedMeld allows patients to pull their medical history from their respective providers in one place
Claimed on Devposthigh confidence
An AI agent derived these features from the project’s Devpost page and readme, then searched the code for each one. Verified features are backed by cited code; claimed-only features had no supporting code, which is not by itself proof a feature is missing.
Export this project's context (description, README, evidence, key source files) to chat with an AI agent elsewhere.