Narasimhan MB

A large-scale infrastructure deployment on AWS that I designed and automated, starting with framework design, tool selection, and environment setup. I implemented best practices for scalable code architecture, incorporating mock tests for quality assurance and early error detection. I also managed cloud security aspects and streamlined infrastructure deployment through automation using GitHub Actions, with optional multiregional deployment support to ensure flexibility and resilience.

As the lead data engineer, I spearheaded a project focused on real-time monitoring of customer shipments to prevent delays. We used Palantir Foundry's Pipeline Builder to build real-time pipelines. I oversaw the entire data lifecycle, from gathering business requirements to implementing scalable, near-real-time data pipelines.

I also developed visualization dashboards that aligned with key business objectives using Workshop and Quiver.

This solution enabled the business to proactively prevent shipment delays while enhancing manufacturing and logistics efficiency through just-in-time production. By reducing unnecessary production, the system optimized operations and improved overall supply chain effectiveness.

Much work during this time required us to extensively use Foundry's Contour, Object Explorer, and Ontology to build and analyze data models and relationships.

As a data engineer, I contributed to a project aimed at analyzing the efficiency of parts production in the automotive sector.

During the analysis, I identified an issue with a key metric used to monitor BOM efficiency: the weight of the parts produced compared to the weight of the theoretical BOM recipe. To investigate further, I focused on several materials with low production efficiency. Through exploratory analysis and reviewing the data distribution leveraging Foundry's Contour and Code workbook, I discovered that the weight of raw materials and components had been recorded in kilograms (kg) instead of grams (g) due to an error in the ERP system's data entry process.

This discovery prompted the implementation of comprehensive data quality checks using Palantir Foundry's Health checks application in the data pipeline and leveraging Foundry's Expectations API at the code level. These measures ensured accurate metrics on the dashboard, as these metrics directly impacted business decisions.