Software Engineer & System Architect. I design, code, and scale distributed systems that power global retail operations. With deep expertise in backend engineering, cloud architecture, and reliability, I bridge the gap between complex code and massive scale.
What I bring to the table as a developer-first leader:
Building high-throughput, low-latency systems using Java, Python, and Node.js.
Architecting resilient microservices, event-driven flows, and cloud-native infrastructure.
Engineering self-healing systems, CI/CD pipelines, and developer tooling.
Integrating LLMs and AI agents to revolutionize observability and operations.
Design principles for massive scale and reliability.
I don't just deploy; I orchestrate. Using Argo and Flagger, I implemented a "Stage1 → Stage4" rollout strategy that validates SLOs (latency, error rates) at every batch (25 → 100 → 500 clusters) before proceeding, preventing fleet-wide outages.
Systems must survive network partitions. I design for the edge, ensuring checkout workflows function even with degraded WAN connectivity, using local caching and eventual consistency models for telemetry.
Deep-dive optimization is key. I use JFR (Java Flight Recorder) to analyze thread pools and GC behavior, tuning heap compaction and eliminating serialization loops to meet strict p99 latency targets.
Driving engineering excellence across the organization.
Pivotal role in certifying 4,600+ clusters for Black Friday. Led full-system stress testing, failure mode validation (packet loss, node churn), and microservice certification to ensure uninterrupted checkout operations.
Engineered a 50% reduction in observability costs by identifying and pruning 200+ unused high-cardinality metrics and optimizing Prometheus scrape intervals.
Reduced Mean Time To Resolution by 40% through the creation of deterministic SRE playbooks and automated root cause analysis templates adopted org-wide.
Technical depth for distributed systems engineering.
Deep dives into complex systems I've architected and built.
Managing 4,600+ independent Kubernetes clusters across retail stores with variable network conditions and hardware. Manual updates were impossible.
Built a GitOps-driven control plane using ArgoCD and Flagger. Designed a custom "Wave Rollout" controller that promotes artifacts from Pilot → Region → National fleet based on real-time health metrics.
Implemented automated canary analysis checking p99 latency and error rates. Built fallback mechanisms for disconnected edge clusters to ensure eventual consistency.
Alert fatigue and slow MTTR due to high volumes of repetitive infrastructure alerts across the massive fleet.
Engineered an autonomous Python agent that listens to alerts, queries Prometheus & OpenObserve for context, and uses an LLM to perform root cause analysis and suggest remediation.
Automated triage for 80% of production alerts and reduced MTTR by 40%, freeing up significant engineering time.
Migrating critical infrastructure components (Helm2→3, Vault→Akeyless) across 4,600+ live clusters without disrupting checkout operations.
Designed dual-read/write paths and fallback logic. Implemented automated linting and dry-run simulations on edge replicas to catch regressions before rollout.
Completed nationwide migrations with zero downtime and 100% data integrity.
Building the tools that build the software.
Automated log parsing and anomaly detection using Python and OpenObserve VRL functions.
Async Python scripts to perform parallel health checks across thousands of K8s clusters.
Integrated JaCoCo and SonarQube into CI pipelines to enforce quality gates automatically.
Exploring the frontier of software engineering.
Designing multi-agent systems that can autonomously manage complex infrastructure and business workflows.
Pushing the limits of concurrency in Java and Python to handle next-generation data volumes.
Building self-driving reliability systems that predict and prevent outages before they impact users.