Sandeep Prajapati

Performance Engineer with 4+ years of experience in load, stress, endurance, and chaos testing for high-traffic, cloud-native applications. Skilled in JMeter, K6, and LoadRunner with expertise across Kubernetes, AWS, Kafka, Elasticsearch, MongoDB, JVM/GC tuning, and full-stack observability (Grafana, SigNoz, ELK, APM). Proven track record identifying bottlenecks, validating scalability and response-time SLAs, and improving system reliability through data-driven tuning and CI/CD-integrated performance engineering.

Performance Engineer I at Tekion Corp (2025-12 – Present)

• Led performance testing for ARC platform modules — including load, stress, and endurance testing — to validate scalability, stability, and response-time SLAs under production-scale workloads.
• Performed end-to-end JVM performance tuning on Java microservices — diagnosing GC pressure (Parallel GC → G1GC), tuning heap allocation (-Xms/-Xmx) and pause-time targets — reducing GC pause times by ~35% and eliminating OOM errors under 2x peak load.
• Executed Kafka consumer performance testing for the Audit service in the PERF environment, validating higher consumer concurrency with stable throughput, zero errors, controlled backlog recovery, and successful scaling under load.
• Optimized Elasticsearch sharding and indexing for a CMS platform supporting 200 enterprise dealers at 20M-customer scale, improving query performance and ES response times with zero API degradation and stable resource utilization under peak load.
• Validated Linkerd service mesh performance across 500 API endpoints, with 89% of endpoints showing <50ms response-time variance and no significant regression during load testing.
• Benchmarked the MongoSupport Library for merged-collection operations, establishing baseline metrics for query performance, throughput, and resource utilization across consolidated MongoDB data-access patterns.
• Ran AWS FIS chaos engineering experiments simulating Redis network failure to validate system resilience and MongoDB fallback capability under production load.
• Developed JMeter and K6 test scripts simulating real-world production workloads to validate service performance and scalability under peak-load conditions.
• Contributed Python automation modules to an in-house tool — adding automated report creation, Elasticsearch data analysis, and JVM diagnostics (thread/heap dump) capabilities — improving operational efficiency by 50%.
• Built performance monitoring dashboards on Grafana, SigNoz, and ES-APM for real-time metrics visualization, including custom Kafka dashboards for consumer lag, throughput, and partition metrics.
• Managed Kubernetes clusters and AWS services for infrastructure provisioning, configuration, and deployment in the PERF environment.

Consultant – Performance Testing at Allstate India (2022-08 – 2025-11)

• Designed and executed end-to-end performance testing strategies for Auto and Property insurance applications, building comprehensive scenarios to validate system performance under varied load conditions.
• Developed and maintained performance test scripts using LoadRunner, simulating real-world user behavior and transaction patterns to ensure application scalability and reliability across peak and sustained load.
• Analyzed test results and collaborated with development teams to identify bottlenecks and recommend tuning improvements ahead of production releases.

B.Tech in Computer Science and Information Technology – Institute of Engineering and Science, IPS Academy (2018 – 2022)