01 / research
Quantitative research
We form hypotheses about how markets behave and test them against large, clean, point-in-time datasets. Only ideas that survive walk-forward and out-of-sample validation become signals.
precipite
Signal, precipitated from noise.
quantitative trading firm · research-driven · fully systematic
Precipite is a quantitative trading firm.
We apply statistics, machine learning and careful engineering to financial markets. Every strategy we trade begins as a research hypothesis, earns its place through rigorous out-of-sample testing, and is executed entirely by software.
Markets are mostly noise. The edges hidden in them are small, short-lived and statistical — invisible to intuition, obvious to a well-posed test. Our work is to let signal settle out of that noise, then trade it with discipline.
01 / observe
Start from raw data.
Returns are noisy, fat-tailed and incomplete. We collect them carefully and look at them honestly.
02 / model
Estimate what's underneath.
Fit the distribution behind the noise, then check that it still holds on data the model has never seen.
03 / decide
Turn probabilities into positions.
Probabilities become position sizes, risk limits and orders, executed by software.
02 / technology
Technology
We build our own research platform, data pipelines and trading infrastructure. Reproducibility, reliability and observability are engineering requirements, not afterthoughts.
03 / trading
Trading & risk
Portfolios are constructed and sized systematically, with risk limits enforced in code. Execution is automated, and the cost of every fill is measured.
- 01No thesis without a test.
Opinions are cheap. Evidence is the bar for every idea, no matter whose it is.
- 02Out-of-sample, or it didn't happen.
A backtest is a hypothesis, not a result. We trust what holds up on data a model has never seen.
- 03Risk first. Return is a residual.
We decide how much we are willing to lose before deciding how much we want to make.
- 04If it can be automated, it is.
People do research and engineering. Machines do the repetitive, the fast and the emotional.
We are researchers and engineers who like hard problems and honest feedback. We hire for curiosity, rigor and ownership — background matters less than how you think.
- quantitative-researcherstatistics, machine learning, and the instinct to try to break your own results
- software-engineerdata infrastructure, distributed systems, production trading code
- quantitative-developerthe bridge between research ideas and systems that trade them
No formal opening needed. Send a CV and something you've built to careers@precipite.com.
- general
- hello@precipite.com
- careers
- careers@precipite.com
type 'help' and press enter.