Machine Learning Quantitative Researcher - Crypto Derivatives

As a Machine Learning Quantitative Researcher at FUKU, you will design, research, prototype, and deploy machine learning models for crypto derivative market-making and systematic trading strategies. You will collaborate closely with quantitative researchers, traders, and low-latency engineering teams to turn statistical and ML insights into production-ready trading systems. This role combines applied machine learning, time-series modelling, market microstructure research, and backtesting to identify alpha, improve quoting performance, optimise risk controls, and enhance execution quality across multiple CEX crypto venues.

What you'll do

• Research, design, prototype, and validate machine learning models for market microstructure, order book dynamics, alpha signal generation, price prediction, spread optimisation, and execution cost modelling for crypto derivatives.
• Build rigorous backtesting, walk-forward validation, and out-of-sample testing pipelines to evaluate ML model performance and conduct robustness checks against overfitting, regime shifts, and exchange-specific quirks.
• Engineer feature sets from high-frequency market data, including order book snapshots, tick data, trade flows, funding rates, volume imbalance, queue dynamics, and latency metrics.
• Partner with traders and quant researchers to translate trading hypotheses into ML frameworks and iterate model design based on live market observations and strategy performance metrics.
• Deploy and integrate trained ML models into low-latency market-making and execution systems, working alongside systems engineers to balance model predictive power with latency constraints.
• Develop monitoring dashboards to track live model performance, feature drift, concept drift, prediction quality, and strategy PnL.
• Perform statistical research on market regimes, liquidity patterns, limit order behaviour, and trading halts to refine model assumptions and risk guardrails.
• Document research findings, model specifications, test methodologies, and production constraints for internal knowledge sharing and compliance review.
• Collaborate on stress testing and scenario simulation to test model behaviour under volatile market conditions, rate limits, and exchange outages.

What you bring

• A Master's or PhD in Computer Science, Statistics, Applied Mathematics, Physics, Quantitative Finance, or a related quantitative field.
• 3+ years of hands-on experience building production machine learning models for systematic trading, market-making, or high-frequency finance in crypto or traditional markets.
• Strong background in time-series ML, including gradient boosting, neural networks, sequence models, probabilistic modelling, anomaly detection, and feature engineering for high-frequency data.
• Proficiency in Python (NumPy, Pandas, Scikit-learn, PyTorch, TensorFlow) and solid statistical foundations including hypothesis testing, probability, Bayesian methods, and correlation analysis.
• Deep understanding of market microstructure: order types, limit order books, funding, liquidity, rate limits, trading halts, and exchange API behaviour.
• Experience with backtesting frameworks, performance attribution, and handling overfitting in financial time series.
• Familiarity with streaming data, event-driven pipelines, Kafka, PubSub, or Redis streams for real-time feature ingestion.
• Strong coding hygiene, experimental rigour, and ability to clearly communicate complex quantitative results to non-ML team members including traders and engineers.
• Knowledge of crypto derivatives, perpetual futures, CEX market behaviour, and exchange edge cases.

Nice to have

• Experience with online learning or reinforcement learning for market making and order execution.
• Hands-on experience with low-latency profiling, performance tuning, and cloud containerised deployments using Docker.
• Publication or open-source contributions related to time-series forecasting or quantitative finance.
• Experience building risk models using ML for real-time position monitoring.
• Experience working with low-latency trading stacks and ability to work with engineers to deploy models into C++ or Rust production systems.

About FUKU and Hytech

FUKU operates within Hytech, a leading management consulting firm headquartered in Australia and Singapore, specialising in digital transformation for fintech and financial services organisations. Hytech applies AI and data-driven approaches to real-world financial use cases including risk management, process optimisation, and decision support. With more than 2,000 professionals worldwide, Hytech has offices across Australia, Singapore, Malaysia, Taiwan, the Philippines, Thailand, Morocco, Cyprus, Dubai, and beyond.