Divya Mounika GURRAM

Drove digital innovation at Amsted Rail through machine learning, delivering measurable cost savings in Operations, Supply-Chain and Engineering departments.

• Developed and deployed production-grade ML pipelines using Azure MLOps for real-time process optimization, supply-chain optimization and predictive maintenance, enabling scalable and automated insights across manufacturing workflows.
• Designed and implemented a predictive model for defect detection and built a prescriptive analytics web app using PTC ThingWorx, reducing product defects by 45% and improving manufacturing yield.
• Partnered with data engineers to architect a unified data model and scalable ETL pipelines, integrating siloed datasets across departments to unlock enterprise-wide analytics.
• Led 13 A/B-style experiments using causal inference methods and custom KPIs at the foundry, resulting in a 10% defect reduction and influencing process changes.
• Collaborated with engineering, IT, and sales teams to gather customer requirements and deliver a machine learning product that directly contributed to customer acquisition.
• Built a deep learning model using RNNs (Keras, Python) to predict force waveforms from acceleration data during railcar impacts, enabling accurate simulations of product behavior.
• Engineered time-series features and developed a cross-validated linear regression model to predict impact velocity with high precision (±0.05 mi/s), improving product testing and design analysis.

• Performed root cause analysis (RCA) on support tickets and used unsupervised learning (clustering) to uncover 10 distinct ticket patterns, reducing triage time and enabling targeted process improvements.
• Led the development of hypothesis-driven dashboards by transforming raw Excel data into interactive Tableau reports, enabling rapid insights through A/B testing and statistical significance analysis.
• Designed and deployed a custom Tableau dashboard integrating 15 high-impact reports, streamlining stakeholder access to KPIs and improving visibility across cross-functional teams.