About Referential Labs

Referential Labs builds observability and robustness tooling for quantitative and algorithmic trading systems. Our focus is the gap between what a backtest promises, what paper trading suggests, and what live execution actually delivers, because that gap is where quant P&L is silently lost.

Why the company exists

Most quant teams discover their infrastructure problems the same way, in production, after they have already cost real money. A backtest reports a Sharpe of 2.1, paper trading looks reasonable, live results disappoint, and nobody can point precisely to why. The instinct is to blame the strategy. Often the strategy is fine and something else is wrong: a data feed went stale for six minutes on Tuesday, corporate actions were not point-in-time adjusted, execution slippage was systematically underestimated in simulation, or a feature quietly leaked future information into training.

These are not exotic failures. They are the routine failure modes of trading infrastructure, and they are largely invisible without deliberate observability. Enterprise TCA vendors frame execution quality as a compliance obligation for large institutions. Open-source backtesting frameworks assume you will build the validation yourself. Neither serves a mid-size quant team that experiences these problems as a P&L problem, not a compliance one. That is the audience we build for.

How we work

The platform surfaces issues for human review rather than making autonomous decisions. In this domain, false positives cost time, false negatives cost money, and the right calibration is specific to your strategies and infrastructure. Automation catches the routine, your team applies judgment to the edge cases, and the platform learns from what you flag.

Beyond the platform, we provide engineering services to help integrate, calibrate, and refine over time. This is not consulting-as-content-marketing; the engagements are limited and the criterion is whether we are the right team to help you with the specific problem in front of you.