1 Layer 1 · Generation
AI generates 2000+ strategies per session.
7 LLM providers, 10 builder pages, 6 strategy categories. Batch generation turns ideation into a factory. Not one strategy at a time, but a population.
Signal Scale Platform
StratCraft
Generate, manage, backtest, and evaluate as many strategies as you can imagine with AI.
Build your own trading system under your control.
Signal Scale Platform · live
Composite equity · IC + ICIR weighted · 1000+ signals
Illustrative: verified in Engine Benchmarks500–1000×
faster vs Python
2000+/sess
strategies generated
7
LLM providers
1000+
signal capacity
The Scale Numbers
What makes institutional-grade signal scale possible for desktop traders.
0
Faster than Python; speed that makes scale practical
0
LLM providers, 2000+ strategies per session
0
Market data sources for backtesting signals
0
Signal capacity in the Alpha Factory
The 3-Layer Competitive Moat
Three compounding advantages that only StratCraft combines. Each one makes the others more powerful.
2 Layer 2 · Speed
C++23 engine, 500–1000× faster than Python.
What makes 1000-signal scale practical. You backtest the entire factory, not just hand-picked favorites. Zero-copy pipelines, deterministic event loop.
Python
(vectorbt)1×
(vectorbt)1×
Rust
(NautilusTrader)42×
(NautilusTrader)42×
StratCraft
C++23784×
C++23784×
3 Layer 3 · Composition
Statistical composition. The institutional layer.
Combine 1000+ signals using IC, ICIR, and Sharpe screening. 5 statistical weighting methods. This is what transforms a backtest library into a signal factory.
∞ All local · no lock-in
Windows, macOS, Linux. All computation runs locally.
No cloud, no API costs at scale. Export portable .py strategies, bring your own LLM keys, and own your research. Zero vendor lock-in.
Windows
macOS
Linux
Why Signal Scale Matters
Retail traders have always been stuck at a structural disadvantage. StratCraft changes the equation.
Conventional Approach
StratCraft Signal Scale
Strategy generation
Hand-craft 3-5 strategies manually
pages.quantnexus.scaleComparison.row1Quantnexus
Backtest throughput
Python: hours per run, test a handful
pages.quantnexus.scaleComparison.row2Quantnexus
Portfolio composition
Pick your best strategy and run it
pages.quantnexus.scaleComparison.row3Quantnexus
Structural approach
Concentration risk. One strategy fails, you fail
pages.quantnexus.scaleComparison.row4Quantnexus
The world's top quantitative funds achieve consistent returns through statistical composition of thousands of signals, not by finding one perfect strategy. StratCraft brings this institutional methodology to your desktop.
The 3-Layer Pipeline
1
Layer 1: Generate at Scale
Describe an idea. The AI Strategy Builder generates hundreds of variants across 6 categories using 7 LLM providers; 2000+ strategies per session, not one at a time.
2
Layer 2: Backtest the Whole Factory
The C++23 engine runs them all, 500-1000x faster than Python. 6+ data sources, live equity curves every 500 bars. Speed is what makes backtesting 1000+ signals viable.
3
Layer 3: Statistical Composition
Bring surviving signals into the Alpha Factory. Apply IC/ICIR/Sharpe screening and combine with 5 statistical weighting methods. The output is a 1000+ signal portfolio. This follows the structural approach used by top quantitative funds.
Start Your Signal Factory
Free tier includes the C++ backtest engine, regime detection, and YFinance + Dukascopy data: everything you need to start building at scale.