Powered by GROK Deep Model

QuantAI 5.0

Intelligent Strategy Generation · Global Data Query · Auto Visualization

More than a backtesting tool. Tell the AI “write an NVIDIA momentum strategy” or “analyze NVDA volatility,” and it auto-writes code, runs backtests, and generates charts.

Core Features Overview Download Windows Client Download Mac Client

Full-Scenario Intelligent Coverage

Powered by the classify_intent intent engine to auto-switch workflows

STRATEGY Mode

Strategy Generation & Backtest

Enter trading logic (e.g., “dual moving average crossover”), and the AI auto-generates strategy class code. Built-in BacktestWorker runs one-click backtests and outputs equity curves.

DATA Mode

Data Query & Visualization

“Plot NVIDIA’s last-year candlestick chart.” The system auto-writes Python scripts calling Matplotlib, embeds charts in the chat, and supports Excel exports.

CHAT Mode

Financial Analysis & Research

Set as a senior CFA analyst persona. Upload data or ask financial questions and the AI delivers deep report analysis with professional insights.

Live Demo: Strategy Backtest & Data Visualization

One window for all quant needs

Intelligent Analysis Chat (PyQt6)
Auto Mode ● Connected
—— 10:05 AM ——
User:
Write a trend momentum strategy for NVIDIA (NVDA). Use 10-day and 30-day crossover, and buy when RSI > 50.
AI Agent: STRATEGY Mode

Got it. Generated NVDA momentum strategy code based on engine_core:

import engine_core as ec
import talib

class NvdaMomentum(ec.Strategy):
    params = (('fast', 10), ('slow', 30), ('rsi_p', 14))

    def init(self):
        # Initialize indicators
        self.ma1 = self.I(talib.SMA, self.data.Close, self.params.fast)
        self.ma2 = self.I(talib.SMA, self.data.Close, self.params.slow)
        self.rsi = self.I(talib.RSI, self.data.Close, self.params.rsi_p)

    def next(self):
        # Core strategy logic
        if not self.position:
            if self.ma1 > self.ma2 and self.rsi > 50:
                self.buy(size=0.95) # Buy with 95% position
        elif self.ma1 < self.ma2:
            self.position.close()

STRATEGY_TO_RUN = 'NvdaMomentum'
Equity Curve (NVDA)
+158.4% Return
—— 10:08 AM ——
User:
Fetch NVDA price data for the last year and plot the closing price trend.
AI Agent: DATA Mode

Got it. Calling Matplotlib to draw NVIDIA price chart:

import pandas as pd
import matplotlib.pyplot as plt

# 1. Fetch data (NVIDIA)
df = engine_core.get_price('NVDA', start='1y')

# 2. Render chart
plt.figure(figsize=(10, 5))
plt.plot(df['Close'], color='#76b900', label='NVDA')
plt.grid(True)

# 3. Render and send
send_to_ai(df)
Script running...
📊 Visualization Result (Matplotlib):
NVDA: $950.02
Vol: 1.25%
Generated by Matplotlib (Agg Backend)
—— 10:12 AM ——
User:
Based on the latest financial data, deeply analyze NVIDIA (NVDA) fundamentals. Is the valuation too high?
AI Agent: CHAT Mode

Called DeepSeek-V3 for online retrieval and analysis of NVDA’s latest financials:

NVIDIA (NVDA) Deep Research Summary
  • 1. Core Growth Engine (Bullish):
    Data center revenue grew 427% YoY, with H100/H200 demand still far exceeding supply. Gross margin remains an astounding 75%+, signaling a wide moat.
  • 2. Valuation Analysis:
    Forward P/E is around 35x. With expected 3-year CAGR above 50%, PEG < 1.0, suggesting it is not significantly overvalued.
  • 3. Risk Notes:
    Watch CSP capex sustainability and AMD MI300 competitive share.
AI Composite Rating:
Strong Buy
Type “Generate PDF Report” and I will format the above into a formal document.
System Log
1. Natural-Language Input

“Write a strategy” or “plot a chart,” and the AI understands instantly.

2. Intelligent Code Generation

Auto-writes Python scripts for strategy logic and data cleaning.

3. Local Engine Execution

Data stays local. BacktestWorker computes at high speed.

4. Visualized Export

Not just charts—one-click Excel report and data exports.