Technical Specialist-Data Engineering

Area(s) of responsibility Data Scientist (Technical Specialist) Specialist Grade Experience: 6–8 years · Location: [To be filled] · Engagement: Full-time Role Expectations • Own the design, development, validation, and productionization of AI/ML models for financial forecasting. • Analyze historical data to identify predictive features and business-relevant patterns. • Develop time-series forecasting models (Prophet, ARIMA, LSTM, XGBoost) tailored to project and quantify forecast accuracy against agreed thresholds. • Build a Conversational AI layer using Azure OpenAI + RAG patterns to enable naturallanguage queries over financial and workforce data. • Partner with FP&A SMEs to translate business drivers into model features and interpretable outputs. Reports To Senior Technical Lead (Python) / Delivery Manager Grade Technical Specialist (above Sr. App Developer, below Sr. Technical Lead) Program FP&A Reporting Automation & AI Forecasting • Deliver explainable AI outputs (SHAP, feature importance, etc.) so that Finance users can trust and act on model recommendations. • Collaborate with the Sr. Application Developer to operationalize models — MLOps pipelines, model registry, monitoring, and retraining triggers. • Document experiments, model cards, assumptions, and limitations for audit and governance review. Must-Have Skills • Python for Data Science: 5+ years hands-on with pandas, NumPy, scikit-learn, statsmodels, and Jupyter-based experimentation. • Time-Series Forecasting: Deep experience with Prophet, ARIMA/SARIMA, exponential smoothing, XGBoost/LightGBM, and deep learning (LSTM/Transformer) for financial or operational forecasting. • Statistical Modeling: Regression, classification, clustering, feature engineering, hyperparameter tuning, cross-validation, and rigorous model evaluation. • Azure ML Stack: Hands-on with Azure Machine Learning (AML), MLflow, Azure OpenAI, and model deployment as endpoints or batch jobs. • LLM & GenAI: Practical experience with prompt engineering, RAG (Retrieval-Augmented Generation), embeddings, vector stores (Azure AI Search / FAISS / Pinecone), and LangChain or Semantic Kernel. • Explainability: SHAP, LIME, partial dependence plots, and communicating model behavior to non-technical Finance stakeholders. Preferred / Good-to-Have Skills • Prior experience with Finance / FP&A / Corporate Planning use cases. • Exposure to Energy, Oil & Gas, or EPC project economics • Exposure to agent frameworks (LangGraph, AutoGen, Semantic Kernel) for building autonomous analytical agents.