Deekobby/Deekobby-ai-

AI research assistant for nurses, researchers, and students in Ghana — web app + WhatsApp bot. # DEEKOBBY AI Research Assistant — Prototype A specialised AI agent to support **nurses, researchers, and students in Ghana and beyond**, with four modes: - 🔬 **Research Assistant** — literature review help, research design, citations, statistics guidance - 🩺 **Clinical Education Q&A** — concept explanations, quizzes/flashcards - 🩹 **Wound Care Support** — educational wound classification & care principles (not diagnostic) - 📢 **Health Promotion Content** — community health messaging drafts See `../docs/DEEKOBBY_AI_Project_Plan.md` for the full project plan, roadmap, and architecture notes. ## Quick Start (Demo Mode — no API key needed) ```bash cd app pip install -r requirements.txt streamlit run app.py ``` Open the URL Streamlit prints (usually localhost). The app works immediately in **Demo Mode**, giving templated structured guidance so you can test the UI/UX and flow before connecting a real AI model. ## Enabling Full AI Responses 1. Copy the example env file: ```bash cp .env.example .env ``` 2. Open `.env` and set: ``` OPENAI_API_KEY=sk-...your-key... ``` (Any OpenAI-compatible provider works — set `OPENAI_BASE_URL` too if using a non-OpenAI endpoint.) 3. Re-run: ```bash streamlit run app.py ``` The sidebar will show "✅ Live AI mode active" once a key is detected. ## Project Structure ``` deekobby-ai/ ├── app/ # Web chat app (Streamlit) │ ├── app.py # Streamlit UI + chat flow │ ├── prompts.py # Mode definitions & system prompts │ ├── llm_client.py # LLM abstraction (live API + offline demo fallback) │ ├── requirements.txt │ └── .env.example ├── whatsapp_bot/ # WhatsApp bot (Flask + Twilio) — see whatsapp_bot/README.md │ ├── bot.py │ ├── prompts.py │ ├── llm_client.py │ ├── requirements.txt │ └── .env.example ├── branding/ │ └── deekobby_logo.png # Logo/brand mark └── docs/ └── DEEKOBBY_AI_Project_Plan.md # Full plan: vision, roadmap, architecture, safety n …