Ai Ml Engineer

Job description

**AI/ML Engineer - Forecasting & Optimization** **Position Overview** We are seeking a Senior AI/ML Engineer to develop and deploy advanced forecasting models and AI-driven features for our Financial Forecasting Platform. You will design and implement closed-loop learning systems that continuously improve forecast accuracy, detect anomalies in real time, and provide intelligent decision support through LLM-powered agents. Your models will directly impact demand planning, margin optimization, and working capital across thousands of product families. **Key Responsibilities** ***Forecasting Model Development*** - Design, train, and deploy advanced demand forecasting models targeting 3-4% MAPE/WAPE accuracy (baseline: 6%) - *MAPE = Mean Absolute Percentage Error; WAPE = Weighted Absolute Percentage Error* - Build models at granular levels (SKU, product family) with scenario analysis capabilities - Implement bias detection and correction for over/under forecast scenarios - Create models that detect price vs. volume trends and mix changes in real time - Develop methods to assess forecast assumption validity and validate assumptions vs. actuals - Compare forecast vs. quote pricing and identify gaps for proactive calibration ***Closed-Loop AI Learning & Continuous Calibration*** - Build systems that automatically test assumptions vs. actuals each forecast cycle - Implement recalibration logic to improve model performance over time - Develop feedback loops to detect when assumptions break and trigger retraining - Create model monitoring and performance tracking systems - Implement A/B testing frameworks to validate model improvements - Design systems that surface model drift and trigger intervention ***Real-Time Anomaly Detection*** - Develop algorithms to detect anomalies in pricing gaps, derates, and forecast deviations - Flag issues such as "price below quote" and overly conservative assumptions - Create early warning systems to surface problems during planning (not post-mortem analysis) - Implement explainability features to help users understand detected anomalies - Design recommendation engines for proactive forecast calibration adjustments ***LLM-Powered Insights & Prompt Engineering*** - Develop prompt engineering strategies for LLM-based financial analysis agents - Build conversational interfaces for complex, multi-source financial questions - Create dynamic, on-demand insight generation capabilities using AI agents - Implement RAG (Retrieval-Augmented Generation) for grounding insights in forecast data - Design systems allowing users to refine and drill into insights through follow-up prompts ***Model Deployment & Operations*** - Deploy models as production APIs serving sub-second latency requirements - Build model versioning and governance systems - Implement automated retraining pipelines - Create documentation and usage guides for model limitations and performance - Support cross-functional teams with result interpretability **Required Qualifications** Education - Bachelor's degree in Computer Science, Engineering, Statistics, Mathematics, or related field - Master's degree in Machine Learning, Data Science, or Operations Research preferred Experience - 7+ years of machine learning development experience - 4+ years building production forecasting or time-series prediction models - 2+ years working with LLMs and prompt engineering - Proven experience improving forecast accuracy metrics (MAPE, WAPE, bias reduction) - Experience deploying models in enterprise environments Technical Skills Programming & ML Frameworks: - Python (primary), R, SQL - Scikit-learn, XGBoost, LightGBM, TensorFlow, PyTorch - ARIMA, Prophet, Exponential Smoothing, LSTM, Transformer-based models LLM & AI: - Experience with GPT, Claude, LLaMA - Prompt design and RAG (Retrieval-Augmented Generation) architectures - Vector databases (Pinecone, FAISS, Weaviate, Milvus) Model Deployment: - MLflow, Weights & Biases, Kubeflow, SageMaker - API serving frameworks (FastAPI, Flask) - Cloud platforms (AWS SageMaker, Azure ML, GCP Vertex AI) Data Processing: - Pandas, PySpark, SQL - Feature engineering at scale - Data validation and quality assurance Domain Knowledge - Strong understanding of demand planning and financial forecasting - Knowledge of manufacturing cost structures and pricing dynamics - Familiarity with working capital and inventory management implications - Understanding of bias, seasonality, and mix changes in forecasting Soft Skills - Strong analytical and problem-solving abilities - Excellent communicationability to explain model performance to non-technical stakeholders - Self-motivated with ability to work independently - Strategic thinking combined with execution excellence - Continuous learning mindset Preferred Qualifications - Experience with causal inference and assumption testing in forecasting - Knowledge of optimization techniques and constraint satisfaction - Familiarity with financial close processes and variance analysis - Experience building explainable AI systems for business users - Publication or patents in forecasting or anomaly detection - Cloud ML platform certifications - Experience with Bayesian methods and uncertainty quantification