Amrit Raj

Used Time Series, XGBoost, and AWS sagemaker to design and develop ML based Predictive model for load forecast, reducing Electric demand forecast error by 20%. Leveraged ML tools, Time Series, and AWS EC2, ECR to develop Predictive model to improve inventory management and reduced stock outs by 25%. Developed CNN based model to identify damaged Electric Smart meters by Image using computer vision being 93% accurate. Leveraged pandas, NumPy, and Feature engineering, enabling precise EDA and models, resulting in a 15% increase in market share. Developed a robust NLP-based model utilizing NLTK library for accurate identification and classification of billing errors within unstructured text data. Collaborated with domain experts and stakeholders to translate business requirements into effective analytical solutions.

Conducted supplier quality audits to ensure supplier product quality meet company standards. Reduced operation time by 25 min/shift and scrap generation by 10% by organizing program nesting. Collaborated with design team in R&D of N.P.D, 625kva and 500kva genset. Led quality improvement initiatives using Six Sigma and 5s methodology resulting in a 20% reduction in defects and a 15% increase in production efficiency. Implemented 11 tank treatment strict quality checks, resulting in a 30% reduction in tank-related defects and a 20% decrease in customer complaints. Trained and mentored new and existing employees in quality control techniques, 5s, FIFO, SPC, and FMEA.