How tushar-81/doc-connect works
Project type and purpose
• DocConnect is a full-stack web application built on Django.
• It combines three Django “apps”—authentication (user management), docAI (automated medical report generation and visualization), and docbot (NLP-powered chatbot)—alongside a small NodeJS service (puppeteer) for PDF/HTML rendering.
• It uses Django templates, Ajax for dynamic updates, Plotly/Dash for charts, PyTorch/NLTK for ML, and Google OAuth2 for secure login.
File structure
• Top-level Django project in “backend/” with asgi.py, wsgi.py, settings.py, urls.py.
• Three Django apps:
• authentication/ – user login/signup, OAuth commands, templates, static CSS
• docAI/ – models for tests, views to generate reports, chart assets, Plotly/Dash code
• docbot/ – chatbot models, training script, NLTK utilities, static/chatbot templates
• Database: sqlite3 with migrations in each app.
• PuppetEE microservice in “puppeteer/” (NodeJS) likely for PDF export.
• scripts/setup.sh for environment setup; requirements.txt for Python dependencies.
README for additional insights
• Tech stack summary: Django, Ajax, Sqlite3, Google OAuth, Dash, Django-allauth, Plotly, PyTorch, NLTK.
• Feature breakdown: automated report generation, interactive visualizations, secure authentication, virtual assistant bot.
Detail without overcomplicating
• Django apps as separate modules.
• NodeJS puppeteer service as a distinct microservice.
• External OAuth2 dependency.
• Static/media file serving (templates, CSS, JS) versus dynamic API calls.
• Don’t omit the ML and visualization libraries—they’re core to docAI and docbot.