Quantitative ML Engineer

As a Quantitative Machine Learning Engineer at Merli, you will help shape the next generation of AI-driven trading infrastructure. This role sits at the intersection of quantitative research, applied ML, and agentic system design, with the goal of optimizing high-frequency (HFT), medium-frequency (MFT), and wholesale trading strategies.You’ll architect adaptive, agent-based ML systems that learn from evolving market microstructures, build robust forecasting and optimization models, and work with trading and infrastructure teams to deploy these solutions in real-world, low-latency environments.Key Responsibilities- Design & Build Agentic ML Systems: Develop autonomous and semi-autonomous agents that perform data acquisition, alpha discovery, backtesting, and execution optimization.- End-to-End ML Engineering: Architect, train, and deploy ML pipelines for HFT/MFT and wholesale trading, from signal generation to execution integration.- Quantitative Research Integration: Collaborate with quant researchers to translate theoretical models into production-ready predictive and optimization systems.- Market Forecasting: Develop deep learning, time series, and reinforcement learning models for price movement prediction and regime detection.- Trading Optimization: Build reinforcement and meta-learning frameworks that adaptively tune strategy parameters in live environments.- Scalable ML Infrastructure: Implement real-time inference, model versioning, and continuous learning pipelines for production systems.- Performance Evaluation: Rigorously validate models with historical and synthetic simulations, ensuring robustness, latency, and financial soundness.- Documentation & Collaboration: Maintain high standards of reproducibility, version control, and code documentation across research and deployment layers.What You’ll Gain- Work at the frontier of AI, quantitative finance, and agentic automation.- Collaborate with quant researchers, data engineers, and trading teams shaping next-gen trading systems.- Exposure to meta-optimization frameworks, reinforcement learning, and multi-agent orchestration.- Hands-on experience with low-latency ML deployment, GPU acceleration, and distributed training in real trading environments.- Ownership of models that directly influence market-making, forecasting, and strategy execution.- Continuous learning and experimentation in a research-first, innovation-driven environment.Qualifications- Bachelor’s, Master’s, or Ph.D. in Computer Science, Applied Mathematics, Financial Engineering, or a related quantitative field.- 3+ years of experience developing and deploying ML models in production (preferably in finance, trading, or large-scale decision systems).- Strong proficiency in Python and ML frameworks: PyTorch, TensorFlow, scikit-learn, NumPy, Pandas.- Deep understanding of supervised, unsupervised, and reinforcement learning, time-series modeling, and probabilistic forecasting.- Experience building scalable data pipelines with Kafka, Flink, or Ray and deploying models in Docker/Kubernetes environments.- Knowledge of market microstructure, portfolio optimization, or signal-based trading systems is highly desirable.- Familiarity with meta-learning, agent-based system design, or multi-agent coordination is a strong plus.- Solid analytical, programming, and debugging skills with an emphasis on system reliability and latency optimization.Preferred Technical Stack- Languages: Python, C++, Rust- ML Infrastructure: Ray, MLflow, Airflow, Weights & Biases- Data Systems: Kafka, Redpanda, Redis, QuestDB- Model Deployment: Triton Inference Server, TorchServe, or custom GPU inference- Cloud/Hybrid Setup: Kubernetes, ArgoCD, Helm