Auto Invest AI

Auto Invest AI is a full-stack stock portfolio analytics platform with:

• user portfolios and default sector portfolios
• a 400-stock StockMaster catalogue
• Bronze/Silver/Gold data pipeline
• precomputed linear-regression forecasts
• fundamentals/news enrichment for table views

Tech Stack

Layer	Technology
Backend	Django, Django REST Framework
Frontend	React (Vite), Tailwind CSS, Recharts
Database	PostgreSQL-ready (configured via Django settings)
Data Sources	yfinance, Economic Times RSS
ML	pandas, scikit-learn (Linear Regression)

Repository Structure

Project-10/
+-- backend/
¦   +-- api/
¦   +-- analytics/
¦   +-- pipeline/
¦   +-- portfolio/
¦   +-- manage.py
+-- frontend/
¦   +-- src/
¦   +-- package.json
+-- docs/

Quick Start

1) Backend

cd backend
conda run -n autosignal pip install -r requirements.txt
conda run -n autosignal python manage.py migrate
conda run -n autosignal python manage.py runserver

2) Frontend

cd frontend
npm install
npm run dev

3) Mandatory data bootstrap (after fresh clone/pull)

cd backend
conda run -n autosignal python manage.py run_pipeline --mode=all

Without this step, portfolio stocks may show Low Data because Gold forecasts and other pipeline outputs are not populated yet.

Current Data Model Notes

• StockMaster is the master stock catalogue.
• PortfolioStock is the through/link table used by portfolio tables.
• Removing a stock from a portfolio removes only the link row from PortfolioStock.
• StockMaster is not deleted when a user removes from portfolio.

Key API Endpoints

Auth

• POST /api/register/
• POST /api/login/

Portfolio

• GET /api/portfolio/
• POST /api/portfolio/
• POST /api/portfolio/{id}/add-stock/
• DELETE /api/portfolio/{id}/remove-stock/?symbol=TICKER

Stocks

• GET /api/portfolio-stocks/?portfolio={id}
• GET /api/stocks/{id}/

Teammate Onboarding Checklist

After cloning or pulling from GitHub:

1. Install backend dependencies.
2. Run migrations.
3. Run the pipeline once (--mode=all).
4. Start backend + frontend.

If pipeline is not run, DB-backed analytics/predictions will appear empty.

Dedicated Pipeline Guide

See: backend/README_PIPELINE.md

Sequence Diagrams

See: docs/SEQUENCE_DIAGRAM.md