03 / 09, 2025
QuantIQ
An Indian equity analysis and event driven backtesting platform with live NSE data and AI research.
Role
Solo engineer, market data, FastAPI backend, React analytics, AI research
Stack
Python · FastAPI · React · NSE API · yfinance · VADER · Groq LLaMA 3.3 70B · WebSocket
Links
The Problem
The README describes QuantIQ India as a professional Indian equity analysis and event driven backtesting platform that works without paid market data API keys.
Indian equity research needs multiple views in one place: live prices, fundamentals, technical indicators, events, backtests, sentiment, and AI generated research that understands Indian market context.
The Architecture
01NSE first data layer
Live quotes and indices come from NSE unofficial APIs with 30 second refresh behavior, while yfinance fills in historical OHLCV, fundamentals, 52 week ranges, market cap, and news.
02Six tab analysis surface
The app organizes analysis into Overview, Technicals, Financials, Events, Backtest, and AI Intel, combining candlesticks, SMA overlays, Bollinger Bands, RSI, MACD, volume, ratios, and event calendars.
03AI Intel and backtesting APIs
FastAPI exposes stock, technical, financial, search, sector, backtest, chat, narrative, research agent, and WebSocket endpoints. Groq powers streaming strategy chat and the fundamentals, news, technicals, and trade thesis research pipeline.
Decisions that mattered
Avoid paid market data by design
The README emphasizes no paid market data keys, so the stack uses NSE and yfinance with graceful partial data fallback rather than requiring institutional feeds.
Model events explicitly
Backtests support quarterly results, RBI policy, Union Budget, and dividends with entry and exit windows, stop loss, take profit, and preset strategy controls.
Use Indian market assumptions
The AI panel understands RBI stance, FII/DII flows, SEBI, NSE/BSE, rupee formatting, Indian fiscal years, and a 6.5% risk free rate for Sharpe and Sortino calculations.
The Numbers
30s
NSE quote refresh
6
analysis tabs
70B
AI research model
0
paid market data keys
What it taught me
A market app earns trust by degrading gracefully when any one source is delayed or partial.
Good financial AI is contextual: the same research workflow becomes more useful when it speaks the market's calendar, rates, symbols, and regulation.