Experience Level: 4–6 Years
Employment Type: Full time
Department: AI Platform / Digital Experience Innovation
Core Tech Stack: Python, XGBoost, Random Forest, Ensemble Methods, Oracle ERP, SQL
Key Responsibilities
Pipeline Development: Design, build, and maintain robust, automated multi-model machine learning pipelines (including XGBoost, Random Forest, and custom Ensemble architectures) for high-accuracy demand and sales forecasting.
Granular Modeling: Develop and fine-tune predictive models at both macro-levels (PPGC) and highly granular levels (SKU) to support localized and global business planning.
Data Engineering & Preprocessing: Architect data preprocessing workflows, handling missing values, scaling, feature engineering, and outlier detection from enterprise data sources.
System Integration: Collaborate with IT and enterprise data teams to orchestrate seamless data integration pipelines between the ML forecasting platform and Oracle ERP systems.
Accuracy & Validation: Continuously measure, track, and optimize model performance using robust metrics,



specifically targeting MAPE (Mean Absolute Percentage Error) and minimizing forecast bias.
Workflow Design: Conceptualize and implement Business Unit (BU) override workflow systems, enabling domain experts to review, validate, and inject manual adjustments safely into automated model outputs.
Required Skills & Qualifications
Experience: 4–6 years of skilled experience as an AI/ML Engineer, Data Scientist, or Predictive Analytics Engineer with a primary focus on time-series forecasting or demand planning.
Core ML Expertise: Advanced, hands-on proficiency in building tree-based models (XGBoost, Random Forest) and implementing Ensemble methods.
Programming: Mastery of Python and its scientific/ML ecosystem (e.g., Pandas, Num Py, Scikit-Learn, Statsmodels).
Enterprise ERP Knowledge: Proven track record of integrating machine learning models or pipeline outputs with Oracle ERP or simil

📌 Ai/ml Engineer Gurugram
🏢 Trigent Software - Professional Services
📍 Gurugram