We are seeking a versatile Machine Learning Scientist / Engineer to design, build, and productionize the algorithms powering our financial intelligence systems. In this hybrid role, you will sit at the intersection of quantitative data science and robust software engineering. You will own the entire lifecycle of our predictive models—from mathematically formulating hypotheses and prototyping advanced models to deploying scalable production pipelines. Your primary focus will be applying ML and time series forecasting to automate Cost Variance, Cost Forecasting, Scenario & What-If Analysis, and KPI Variance
Key Responsibilities
- Advanced Predictive Modeling: Design, train, and validate sophisticated machine learning architectures and classical statistical models tailored for multi-horizon cost forecasting and KPI predictions.
- Time Series & Sequential Modeling: Leverage advanced time series techniques (e.g., Deep Learning, State-Space models, hierarchical forecasting) to capture complex seasonal patterns, macroeconomic dependencies, and trend shifts in high-dimensional financial data.
- Scenario & What-If Simulation: Develop simulation engines (such as Monte Carlo and stress-testing frameworks) that allow financial planners to run interactive What-If scenarios, modeling the ripple effect of operational and market changes on cost structures.
- KPI & Cost Variance Analysis: Build automated anomaly detection and diagnostic models to pinpoint the root causes of variance between planned, forecasted, and actual financial KPIs.
- Production Pipeline & MLOps Engineering: Refactor prototype code into clean, scalable production services. Deploy and containerize models, orchestrate pipelines, and build monitoring systems to detect feature and model drift over time.
- Financial Translation: Partner with corporate finance teams to translate complex statistical outputs into transparent, interpretable insights and interactive strategic dashboards.
Data Science & Modeling Expertise
- ML & Statistical Foundations: Strong theoretical and practical foundation in supervised/unsupervised learning, probabilistic programming, ensemble methods, and non-linear regression.
- Deep Time Series Domain: Extensive experience with forecasting frameworks (e.g., Prophet, ARIMA, DeepAR, Temporal Fusion Transformers, or N-BEATS) and handling sparse, noisy, or irregular financial datasets.
- Simulation & Decision Science: Proven ability to build simulation frameworks, sensitivity analyses, or Bayesian networks for risk and scenario modeling.
Software & MLOps Engineering
- Core Tech Stack: Mastery of Python and its scientific/ML stack (Pandas, NumPy, Scikit-Learn, PyTorch/TensorFlow, or JAX).
- Engineering & Scale: Strong software engineering practices (Git, unit testing, APIs) with experience scaling computations using distributed frameworks (e.g., Spark, Ray) for heavy simulation workloads.
- Data & Cloud Systems: Proficiency in SQL and cloud data warehouses (e.g., Snowflake, BigQuery) alongside MLOps orchestration tools (e.g., Docker, MLflow, Airflow, or Kubernetes).
Experience & Education
- Education: Master's or Ph.D. in Data Science, Computer Science, Statistics, Quantitative Finance, or a highly quantitative field.
- Experience: 5+ years of professional experience as a Data Scientist or MLE. Nice to have would be a clear history of applying machine learning directly to financial, economic, or operational planning data.
- Domain Knowledge: It will be very good if he/she has a solid grasp of corporate finance principles (budgeting cycles, driver-based planning, cost allocation, and variance attribution).
