ROLE-004 · TRADING · 2026

Algorithmic Trader

Research, deploy, and refine systematic strategies across crypto and equities with a strong bias toward execution discipline, market structure awareness, and rigorous validation.

Type[ Full-time ]·Discipline[ TRADING ]

The role

This is not a pundit role and it is not pure paper research. Gordon starts with agentic trading across crypto and stocks because markets are the fastest proving ground for delegated financial action: outcomes are measurable, feedback loops are tight, and bad assumptions get punished quickly. We want someone who treats trading as a systems problem. The job spans signal research, backtesting, portfolio construction, execution logic, live monitoring, and the operational discipline required to keep real strategies honest once they leave research mode. You should care as much about costs, liquidity, latency, and failure modes as about alpha. We want traders who can move between hypothesis and implementation, collaborate tightly with engineers, and know when a model is genuinely robust versus merely attractive in a backtest. A lot of the leverage in this role comes from closing the loop between research and reality: understanding when a strategy is drifting, when market structure changed underneath you, when the right answer is to tighten risk, and when the right answer is to kill a strategy entirely. Over time, this role also helps define what responsible, delegated trading should look like inside Gordon itself.

What You'll Build

Systematic strategies across supported crypto and equity markets where we believe we can earn the right to compete.
Research and simulation workflows that incorporate fees, slippage, liquidity constraints, regime changes, and realistic execution assumptions rather than fantasy fills.
Forecasting, sequence-modeling, and foundation-model experiments for market data where they materially improve signal quality, regime detection, or decision support rather than just adding sophistication.
Execution logic, sizing rules, and portfolio constraints that survive contact with live market conditions.
Monitoring, alerting, drift detection, and post-trade review systems that keep live behavior legible and actionable.
Risk frameworks for exposure, concentration, venue behavior, and failure scenarios that matter in real trading rather than just in research notebooks.
A tighter bridge between strategy research and runtime behavior so we can tell quickly whether live performance confirms or invalidates the thesis.
The trading logic that eventually informs what Gordon should do, what it should not do, and how much confidence it should have when acting on behalf of a user.

How We Think

Trading is a systems problem. Edge only matters if it survives implementation, execution, and risk.
Backtests are evidence, not truth. A strategy is only interesting if it survives realistic assumptions and then survives the market.
Market structure matters. Queue position, liquidity, spread behavior, order types, session structure, venue quality, and the behavior of other participants are not implementation details.
Net performance matters more than elegant theory. Fees, slippage, funding, borrow, latency, and operational drag are part of the strategy.
Risk management is part of alpha. A strategy that cannot be monitored, sized responsibly, or shut down cleanly is not a finished strategy.
We care about repeatable process over narrative conviction. The goal is not to defend a view; it is to discover what holds up.
We would rather kill a weak strategy early than carry intellectual sunk cost into live losses.
The work is collaborative by nature. Serious trading sits at the intersection of trader judgment, quantitative research, and engineering quality.
Crypto and equities are a useful proving ground because they are liquid, information-rich, and already wired into programmable execution environments.
Automation raises the bar for discipline. The easier it becomes to launch strategies, the more important validation, monitoring, and operational restraint become.

How You Work

You start with hypotheses, data, and market behavior, then work toward deployable strategy logic instead of stopping at research output.
You are comfortable coding your own research, validating assumptions aggressively, and working with engineers to productionize what deserves to go live.
You know how to separate signal from artifact. You look for data leakage, overfitting, survivorship bias, unstable parameter sensitivity, and hidden execution assumptions before trusting results.
You are comfortable reasoning probabilistically about markets: distributions, regimes, uncertainty, conditional forecasts, and how confidence should affect action or abstention.
You monitor live behavior closely and treat deviations from expectation as a prompt for investigation, not rationalization.
You can explain a strategy clearly: what edge it is trying to capture, what conditions it depends on, what breaks it, and how we would know.
You are comfortable working across research notebooks, data pipelines, dashboards, logs, and live trading controls in the same week.
You can operate in both fast loops and patient loops: some problems need intraday reaction, others need slow careful validation.
You respect the operational side of trading. Trade reconciliation, venue quirks, parameter hygiene, and post-trade analysis are part of the job, not admin around the job.
You are self-directed enough to identify the next strategy question, the next weakness in the stack, or the next risk issue without waiting for a perfectly specified prompt.
You are comfortable using modern tooling, including AI-assisted analysis and coding, but you do not outsource market judgment or model validation to a model.

Requirements

Clear evidence that you have researched, deployed, or operated systematic trading strategies in real markets or in rigorous trading environments.
Strong quantitative ability across statistics, probability, empirical analysis, and the habit of reducing trading questions to testable assumptions.
Strong knowledge of market microstructure, execution, and the difference between theoretical edge and tradable edge.
Strong programming ability for research and strategy development. Python is expected; deeper systems fluency or close collaboration with low-latency engineers is a strong plus.
Capability in modern predictive modeling for markets, including probabilistic time-series forecasting, sequence modeling, or training and adapting foundation-style models on financial data, is a strong plus.
Experience building or using backtesting, simulation, and monitoring workflows that account for realistic costs and operational constraints.
Comfort with portfolio construction, position sizing, risk limits, and the practical tradeoffs between aggressiveness and survivability.
Ability to work with large volumes of historical and real-time market data and maintain a high bar for data quality.
Comfort in crypto, equities, or both. Cross-asset curiosity is valuable because Gordon's initial wedge spans both worlds.
Strong judgment under uncertainty. You should know when to push, when to reduce size, when to wait, and when to stop.
Good communication and collaboration skills. We need people who can work tightly with engineers, explain assumptions, and improve the system as a team sport.
Respect for operational rigor. Live strategies need clear controls, observability, review, and postmortem discipline.
A bias toward process over ego. The market does not care about elegance, and neither should we when the evidence says otherwise.

Example Problems

Take a promising research result and determine whether it still works after realistic assumptions for fees, spread capture, slippage, latency, and liquidity.
Build or refine a strategy that trades across crypto and equity instruments, then explain exactly where the edge comes from and under what regimes it should or should not run.
Diagnose why a live strategy deviated from backtest expectations and determine whether the issue is data quality, execution, market structure drift, risk settings, or a broken thesis.
Design a monitoring and kill-switch framework that makes it obvious when a strategy should be paused, resized, or retired.
Translate a human market intuition into a testable systematic rule, then decide honestly whether it survives contact with evidence.
Evaluate whether a probabilistic forecasting stack or a fine-tuned foundation model on market data actually adds incremental edge over simpler baselines once costs, latency, and robustness are accounted for.
Improve execution quality on a strategy where raw signal may be fine but net performance is getting destroyed by impact, spread crossing, or poor venue behavior.
Build a post-trade review process that helps us understand whether losses came from expected variance, hidden concentration, operational mistakes, or model failure.
Figure out how a delegated trading system should express confidence, abstain from action, or size down when market conditions become ambiguous or hostile.
Work with engineering to make a strategy safer and more observable without slowing iteration to a halt.
Know when the highest-leverage move is not to add another model, but to improve validation, execution, or risk controls on the strategies we already trust.

Apply

Tell us why this role is the one.

Send a short note and your best evidence of the work to contact@generalliquidity.com. Every position begins with a short paid trial period.

Apply for this role→Back to open roles→