Quantitative Lead – Client Profiling & Trading Intelligence

Role purpose

Lead the use of quantitative analysis and AI to understand client trading behaviour, execution outcomes and the brokerage’s exposure to different types of flow. Define a robust client profiling and trading intelligence methodology and turn analytical findings into explainable recommendations for Dealing, Risk and senior management. Combine market knowledge with hands-on research and leadership of quantitative model development.

Key responsibilities

Profiling strategy and research leadership

• Define a client and flow profiling framework using trading frequency, holding periods, instrument mix, position sizing, leverage, timing and execution behaviour.
• Develop AI-led segmentation, clustering, anomaly detection and predictive research priorities, with interpretable statistical baselines and clearly defined business questions.
• Distinguish observed trading behaviour, economic exposure and operational anomalies; avoid treating short-term profitability or a model-generated label as proof of abusive activity.
• Lead the Quantitative AI/ML Engineer’s research backlog, review feature and label definitions and establish standards for experiments, documentation and reproducibility.

Trading intelligence and business application

• Analyse flow characteristics, mark-outs, slippage, concentration and correlated activity to understand execution quality, adverse selection and changes in market conditions.
• Assess broker exposure and hedging outcomes under the approved execution model, separating client trading outcomes, broker revenue, hedge costs and residual market risk.
• Develop dashboards, cohort analyses and decision-support reports that explain the drivers of changes in client behaviour and trading risk.
• Present evidence-based recommendations to Dealing and Risk, including confidence, alternatives and potential client impacts; route implementation through approved decision processes.

Model governance and team leadership

• Review out-of-time performance, label quality, regime sensitivity, fairness where relevant and the economic significance of model results before recommending operational use.
• Define monitoring thresholds, review schedules and escalation triggers for profiles and trading intelligence models; ensure profiles can change as new evidence emerges.
• Partner with Compliance on permitted data use and client treatment, and with independent reviewers on material model challenge and validation.
• Coach quantitative staff, maintain a research roadmap and communicate findings to senior leadership without overstating model certainty or implying guaranteed trading performance.

Essential experience and qualifications

• 8+ years in quantitative research, trading analytics, electronic trading or market risk, including 3+ years leading research initiatives or quantitative staff.
• Direct FX/CFD brokerage experience with a strong understanding of retail client flow, broker execution models, hedging, liquidity and trading risk.
• Strong hands-on Python and SQL, statistical modelling and applied machine-learning skills; ability to review models and investigate results independently.
• Experience building behavioural profiles or trading intelligence using transaction-level data and translating findings into operational decision support.
• Deep understanding of validation pitfalls, including leakage, selection bias, regime changes and misleading results from limited trading histories.
• Ability to explain technical findings clearly to Dealing, Risk and executive management and challenge commercially attractive but weakly supported conclusions.
• Degree in a quantitative discipline; a master’s degree or PhD is advantageous but does not replace relevant brokerage experience.

Preferred experience

• Experience establishing a quantitative research function or moving profiling models from research into monitored production.
• Experience with MT4/MT5 datasets, liquidity-provider analytics, execution mark-outs and broker risk-book analysis.
• Experience with graph analysis, behavioural clustering, explainable ML and controlled challenger assessments.