Aman Thakkar

Just joined Meta as a Software Performance and Capacity Engineer. Please reach out to me if you'd like to know more!

I worked on building the cloud platform for the next generation of Aruba's (a Hewlett Packard Enterprise company) enterprise switches.

• Detected & resolved race conditions in concurrent goroutines by redesigning high-volume message processing pipeline
• Eliminated OOM errors in Kubernetes pods with efficient memory usage along with distributed message tracing and logging
• Achieved ~30% lower inter-service communication latency by transitioning multiple legacy APIs from REST to gRPC
• Optimized Golang cloud service for Aruba switches with Redis caching, lowering database queries by 80%
• Boosted feature branch testing speed by ~10% across organization by improving Python-based test automation framework
• Led effort to deliver multiple workflows and services for Aruba's on-premise solution with intensive testing within tight deadlines
• Enhanced real-time service performance insights via Prometheus metric collection and custom Grafana dashboards
• Developed Proof of Concept application using RAG techniques to provide automated Level 1 triages for bugs based on runbooks from multiple services

Glidewell Dental, based in Irvine, California, is a leading dental laboratory known for its innovative use of technology to enhance the quality and efficiency of its services.

• Re-designed front-end architecture of mobile app by independently leading migration from Angular to React Native
• Achieved 50% reduction in crown validation time by integrating 3D renders within the app using Three.js
• Minimized backend server load through real-time communication with milling devices with advanced web socket implementation
• Slashed hardware bug reporting time by 90% with new feature allowing dentists to snap and submit photos with seconds

Barclays Global Service Center (BGSC) is a global capability center of Barclays Bank PLCI worked with Payment Sanctions in the Financial Crimes Division, on multiple projects and internal services.

• Saved USD 30,000 in capital costs by developing an in-house website using React for Operations to flag fraudulent transactions
• Led cross functional effort to develop, test and deploy new payment route configurations for real-time, cross-border euro transfers
• Reduced manual effort by ~75% by creating Python scripts to detect live data breaches in test environments
• Boosted analysis efficiency by ~30% with Python parser to process raw payment files of over 50 types into customizable reports
• Achieved 5x increase in code coverage through the utilization of Selenium and Cucumber for test case automation