Employment type: Contract

Industry: Banking

Area: IT

Location: Zurich

Remote from abroad?: No

Work load: 40%

Tasks and responsibilities

• Support the design, development, and roll-out of a global market abuse detection model for the client
• Work with Compliance Analytics and Transformation to translate regulatory and business requirements into data science use cases and model specifications
• Explore, clean, and preprocess complex financial time series data, including irregular and microstructural data (jumps, liquidity gaps, roll adjustments, observational bias)
• Design and implement time-series–based outlier and anomaly detection logic to identify abnormal market behaviour
• Link detected anomalies to external events such as macroeconomic releases, central bank decisions, and corporate actions; integrate NLP-based news or sentiment signals where relevant
• Build, validate, and document statistical and machine learning models (including volatility and econometric models) using Python and relevant libraries
• Implement robust, production-grade Python components (modular design, exception handling, performance profiling, tests) and reusable internal libraries
• Write and optimise complex SQL queries (primarily Oracle) over large-scale datasets
• Own the full model lifecycle: requirements gathering, feature engineering, model development, validation, deployment, monitoring, documentation, and maintenance
• Challenge assumptions, ensure analytical rigor, and prevent issues such as spurious correlations, lookahead bias, and flawed experimental design
• Collaborate with cross-functional stakeholders, explain model behaviour and findings, and turn technical results into actionable insights for non-technical audiences

Must-have criteria

• At least 7 years of professional experience in quantitative roles in top-tier banks, asset managers, or fintechs in capital markets
• Strong exposure to regulatorily sensitive domains such as market or trade surveillance, market abuse monitoring, or comparable compliance / risk areas
• Master’s degree or PhD in Quantitative Finance, Mathematics, Physics, Engineering, or a closely related quantitative field
• Strong SQL skills, ideally with large-scale Oracle environments
• Demonstrated end-to-end ownership of analytical models (from requirements to monitoring and maintenance)
• Evidence of high analytical rigor and critical thinking (e.g. challenging assumptions, improving model robustness, preventing false conclusions)
• Excellent communication skills and ability to explain complex models and findings to non-technical stakeholders
• Expert Python skills for data science, including:
• Object-oriented programming and modular architecture
• Exception handling, performance profiling, and test-driven development
• Deep hands-on use of pandas, NumPy, SciPy, scikit-learn, statsmodels, and at least one of PyTorch or TensorFlow
• Extensive experience working with financial time series data and microstructural challenges (irregular and asynchronous data, jumps, liquidity gaps, roll adjustments, observational bias)
• Proven track record in designing and operationalising time-series anomaly or outlier detection systems for abnormal market behaviour
• Advanced knowledge of time-series econometrics and anomaly detection, including:
• ARIMA / SARIMA / VAR
• GARCH-family models (e.g. EGARCH, GJR-GARCH)
• Change-point detection (e.g. PELT, Binary Segmentation)
• Volatility-adjusted residual and forecast error analysis
• Unsupervised methods (e.g. Isolation Forest, One-Class SVM, DBSCAN, LSTM autoencoders)

Nice-to-have criteria

• Direct experience building models specifically for market abuse, trade surveillance, or similar regulatory use cases
• Hands-on experience integrating NLP / news sentiment feeds into event-driven detection logic
• Prior work in global banks with complex, multi-market data landscapes
• Experience working closely with Compliance, Risk, and Front Office stakeholders on surveillance and monitoring topics
• Familiarity with MLOps practices and tools for monitoring model performance and drift in production
• Experience contributing to internal Python packages or analytics libraries in a financial institution

Contract duration

• 3 months with option of extension

Language requirements

• English