Headline

Performance Engineering Leader | Platform Architect | SRE Strategist | Cloud-Native Scalability | Observability Architecture | Building Scalable Systems & High-Performing Teams

About

Performance Engineering Leader and Platform Architect with 18+ years designing high-throughput distributed systems, performance frameworks, and cloud-native architectures for enterprise SaaS at scale. What I do: • Architect performance testing frameworks, observability platforms, and infrastructure recommendation engines that operate at enterprise scale • Lead cloud-native migrations (EC2 to Kubernetes) with predictive auto-scaling, saving millions in annual infrastructure spend • Build automation that eliminates manual bottlenecks, including a post-test analysis pipeline that reduced engineering analysis time by 83% • Design and benchmark microservices (Identity, Central Messaging) with P50/P95/P99 profiling and capacity planning • Mentor and grow engineering teams, scaling a performance org from 2 to 18 engineers with 3 promoted to senior roles Notable work: • Co-designed a patented Performance Benchmarking Service (PBS) with an infrastructure recommendation engine that analyzes production deployments and recommends EC2 right-sizing, saving millions across 200+ customer environments • Resolved a critical EC2-to-Kubernetes migration blocker by switching load balancing from Round Robin to Least Request, reducing P95 queue wait time by ~80% • Built "Dyno," a Dockerized self-service regression suite that shifts performance testing left, reducing post-merge incidents by ~40% • Engineered end-to-end automated workflows (Sentinel/n8n) from lab provisioning to workload execution to results analysis • Built AI-powered performance skills using Claude Code for automated code analysis, Playwright-based script generation, and N+1/caching anti-pattern detection I speak at conferences (RubyConf), participate in architecture design reviews for major features, and champion hackathon-driven innovation. Currently focused on Kubernetes migration strategies, predictive scaling, and observability-driven cost optimization. Tools & Technologies: Gatling, JMeter, LoadRunner, k6, Playwright, AWS (EC2, ECS, EKS, RDS), Kubernetes, Docker, New Relic (NRQL, APM, Infra, API), Datadog, Dynatrace, Grafana, ELK, Jenkins, n8n, Claude Code, Python, Scala, Java, Ruby on Rails, MySQL, Redis, Elasticsearch, MongoDB Open to conversations about senior technical leadership, platform architecture, and SRE strategy roles.

Location

Pune Division

AI stack

anthropic apis

claude code

claude

cursor

gemini

github copilot

Work experience

Lead Performance Engineer at Coupa Software, Nov 2017 - Present · 8 yrs 10 mos, Pune District, Architected a modular Gatling-based performance framework from scratch with parallel multi-instance execution and CI/CD integration, expanding coverage by 3x while reducing maintenance effort by ~70%. Co-designed the Performance Benchmarking Service (PBS), a patented platform profiling customer transaction flows and identifying latency hotspots; deployed to 100+ enterprise customers. Enhanced PBS with an infrastructure recommendation engine analyzing production telemetry for EC2 right-sizing, saving millions annually across 200+ environments. Debugged a critical EC2-to-Kubernetes migration blocker caused by passenger queuing, switched load balancing from Round Robin to Least Request, reducing P95 queue wait time by ~80% and unblocking the migration org-wide. Built “Dyno,” a Dockerized self-service regression suite for developers to validate performance on feature branches before merge, reducing post-merge incidents by ~40%. Engineered automated post-test analysis using New Relic API and Google Apps Script, reducing per-engineer analysis from 60 to 10 minutes (83% reduction), reclaiming ~100+ engineering hours monthly. Designed end-to-end automated workflows (Sentinel/n8n) from lab provisioning to results analysis. Built AI-powered skills using Claude Code for code analysis, Playwright script generation, and N+1/caching anti-pattern scanning. Spearheaded Resque jobs and frontend server migration to Kubernetes, engineered predictive auto-scaling based on historical traffic, achieving ~35% cost reduction while handling 2x traffic spikes. Established release validation gates for major upgrades (Elasticsearch, Redis, Ruby, Rails), catching 12+ regressions pre-production. Ran daily workload suites with root cause profiling, reducing repeat incidents by ~30%. Grew the India performance team from 2 to 18 engineers, 3 promoted to senior roles. Presented at RubyConf and Engineering All Hands. Championed hackathon culture, delivering POCs that shipped to production.

Senior Performance Engineer at IBM India Private Limited, Jul 2016 - Nov 2017 · 1 yr 5 mos, Pune, Maharashtra, India, Engineered comprehensive JMeter-based load tests for the Emptoris suite (Spend Analysis, Sourcing, Contract Management), profiled Oracle/MySQL/DB2 query execution plans and optimized slow queries, connection pooling, and indexing, improving throughput by ~25% and reducing query time by 40%. Automated the end-to-end performance suite with ELK Stack, Jenkins, and Maven, enabling scheduled unattended execution with automated dashboards, reducing test cycle time from 3–4 days to under 1 day.

Specialist - Performance Testing at SunGard - now part of FIS, Jul 2010 - Jun 2016 · 6 yrs, Pune/Pimpri-Chinchwad Area, Optimized deployment processes using Tomcat, WebLogic, ANT, Maven, and Shell scripting, implemented Jenkins CI/CD pipelines for automated build and deploy, reducing deployment time by 40% (5 hrs to 3 hrs) and decreasing deployment incidents by ~50%. Engineered automated test scripts using LoadUI/SoapUI and JMeter for SOAP/REST web services validation during HTML5 framework migration, ensuring functional and performance correctness across 50+ API endpoints.

Software Engineer at Cognizant Technology Solutions, Dec 2007 - May 2010 · 2 yrs 6 mos, Pune/Pimpri-Chinchwad Area, Designed and developed enterprise software solutions built automated test frameworks that reduced manual testing effort by ~60% and accelerated delivery cycles across 3 client engagements.

Education

Bachelor's Degree, Computer Sciences (B.C.S), Sant Gadge Baba Amravati University, Amravati

Master's degree, Information Technology, Symbiosis Center of Distance Learning

Public links

github: https://github.com/pankajdhoot