Where Noise Ends,
Alpha Begins.
AlphaEnd Labs applies large language models to quantitative factor discovery — searching wide, selecting hard, and keeping only the signal that survives out of sample.
We mine the signal that survives the noise.
AlphaEnd Labs is a ten-person research lab building the systems that turn raw market data into deployable alpha. Today that work centers on LLM-driven factor discovery — generating, mutating, and crossing over hypotheses under structured quality gates, then distilling them into a compact library of low-correlation alphas, evaluated with the discipline of a trading desk.
We think quant should be more open — the field advances when the tools are shared, even if the alphas aren't. So we open-source the method, not the results: run the framework and mine your own.
What we build.
alpha-loop
An open-source, LLM-driven factor-mining framework — diversified parallel search, trajectory-level evolution, and structured quality gates. Bring your own model, universe, and data, and mine your own alphas.
Run the framework →stock-benchmark
A benchmark that pits time-series architectures against one another as alpha combiners — turning a factor library into a single prediction, then a cost-aware portfolio.
View repository →On the roadmap — neutralization, portfolio construction, high-frequency signals, and end-to-end execution.
Method, not hype.
Breadth
Any single idea has a low chance of working — that's the base rate in quant research. The edge isn't a cleverer idea; it's covering more of the hypothesis space in the time you have.
Discipline
More ideas, more noise. A strong backtest is table stakes — the signal still has to hold across other periods and other universes, and not just restate something you already own.
Out-of-sample
The prettiest backtests are usually the most-tried ones. What you look at while researching and what you judge on must stay apart — or you're reporting overfitting, not return.