Skip to content

kolbis/takecare-ai

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TakeCare — AI-Powered Medication Assistant (WhatsApp)

⚠️ Work in Progress This repository is under active development. Architecture, APIs, and features may change frequently.

TakeCare is an AI-powered medication assistant built on WhatsApp. It helps older adults manage complex medication schedules safely and reliably, while keeping caregivers informed when needed.


🌿 Motivation

This project was born from a personal need.

My mom was recently diagnosed with Acute Myeloid Leukemia (AML). She now takes more than 20 different pills every day, often while feeling overwhelmed or confused, and she lives far away.

TakeCare is being built to support her — and families like ours — by providing:

  • Clear, simple medication guidance
  • Timely reminders
  • Safety checks
  • Caregiver notifications when something seems wrong

This is not just a technical experiment. It’s a real-world tool solving a real-world problem.


🏗 Architecture Overview

The project follows a DDD-inspired modular structure.

.
├── api/
├── app/
├── agentic/
└── shared/

api/

FastAPI layer:

  • WhatsApp webhooks
  • REST endpoints
  • Scheduler
  • Caregiver notifications
  • Infrastructure adapters

app/

Core application layer (DDD):

  • Use cases
  • Domain entities & value objects
  • Repository interfaces (currently mocked)
  • RAG orchestration

agentic/

AI orchestration layer:

  • LangGraph flows
  • Tools
  • Skills
  • Prompts
  • Agent graph definitions

shared/

Cross-cutting components:

  • Shared types
  • i18n (EN / HE)
  • Utilities

🚧 Current Status

  • Core domain modeling in progress
  • Agent workflows under active experimentation
  • Infrastructure adapters evolving
  • Database integration not finalized
  • APIs may change without notice

This repository should be considered unstable and experimental at this stage.


⚙️ Setup

Install dependencies:

uv sync

Run the API:

uv run --project api uvicorn api.main:app --reload

Run a specific package:

uv run --project agentic python -c "from agentic.graph.graph import graph; print(graph)"

🎯 Vision

TakeCare aims to become:

  • A reliable AI medication companion
  • A safety layer for complex treatments
  • A bridge between patients and caregivers
  • A multilingual (EN/HE) assistant accessible through familiar messaging platforms

About

AI-powered medication adherence assistant for elderly users, built on LangGraph and DeepAgents with safety-first orchestration.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages