Theme color
Display
Press ? for shortcuts · ⌘K for command palette
OPERATING THE PLATFORM BEHIND 165+ FORTUNE-500 SAAS TENANTS
I keep the platform running for 165+ Fortune-500 customer environments. When something breaks at 3 a.m., I'm the one writing the RCA.
Bengaluru, India · Open to relocation
It's --:-- for me right now
Last RCA shipped 3d ago · Last deploy reviewed today · Bytecode mood calm
01 — Impact
xargs -P across 14 envs at oncejavap · 13-section RCA02 — About
I'm an SRE / DevOps engineer at Hyland, where I run multi-tenant SaaS on AWS for 165+ Fortune-500 customer environments alongside 32 Nuxeo Cloud tenants. The work spans multiple AWS accounts and regions: 24×7 incident response on Datadog + PagerDuty, blue-green PTM upgrades on EKS, Terraform-driven infrastructure, JVM-level RCAs, and platform automation in production.
5+ years in SRE / DevOps (6+ in tech with earlier software-engineering internships), supporting customers in financial services, automotive, consumer goods, and insurance. I've also operated workloads on Azure (AKS) and GCP (GKE) across customer accounts.
03 — Skills
Four capabilities I'm hired for. The full tool list lives below, collapsed by default.
Running Alfresco + Nuxeo SaaS for Fortune-500 customers across 165+ AWS environments and 32 Nuxeo tenants. Blue-green PTM upgrades on Kubernetes, Solr shard-lag gating, ASG topology and ALB session stickiness, Hazelcast cluster-split recovery.
Stack: AWS · EKS · Terraform · Aurora · Solr · Hazelcast · ActiveMQ · Kafka (MSK)
Bytecode-level RCAs with javap through Spring transaction internals. Aurora PostgreSQL / MySQL EXPLAIN tuning, parameter-group hardening, blue-green engine upgrades, plus a parallel diagnostic that cut fleet triage from ~80 m to 1 m 34 s.
Stack: Java · javap · Spring · Aurora · pgBadger · pg_stat_statements · Python
Owner of the 24×7 PagerDuty rotation. Datadog monitors, dashboards, and SLOs for Tomcat threads, DB ACUs, Solr remaining-TX, customer exception tracking, and ALB IP rotation. Word-format RCAs with reproducible evidence, runbook authoring, blast-radius control.
Stack: Datadog · PagerDuty · OpenTelemetry · Grafana · Splunk · CloudWatch
Investigation pipelines that classify the entry path and correlate evidence across PagerDuty, Datadog, Jira, AWS, EKS, DB, and Solr. Three-tier guardrails (auto-allow / ask / deny) with secret redaction, PII exfil detection, and audit logging, built on 9 cooperating Python and Bash hooks.
Stack: Terraform · Python · Bash · GitHub Actions · Jenkins · MCP · Claude Code
EC2, EKS, ECS, Fargate, Lambda, RDS / Aurora (PostgreSQL, MySQL), Aurora Serverless v2, Aurora DSQL, DynamoDB, ElastiCache, S3, EBS, ECR, IAM, KMS, SSM, VPC, ALB / NLB, Route 53, API Gateway, Auto Scaling, Step Functions, EventBridge, SNS, SQS, MSK, AWS MQ, Athena, Glue, Kinesis, Firehose, CloudWatch, CloudTrail, AWS Backup, AWS Config, AWS Organizations, Control Tower, Inspector, Secrets Manager, WAF, GuardDuty, Security Hub, Cross-Account IAM, Cost Explorer, Compute Optimizer, Trusted Advisor.
Claude Code, ChatGPT (Codex), GitHub Copilot, Google Gemini, DeepSeek, Perplexity, AWS Bedrock, MCP (Model Context Protocol), Hugging Face — AI-driven RCA generation, hook-based agent guardrails, MCP integrations with Datadog / PagerDuty / Atlassian / PaaS, prompt-engineered investigation pipelines.
Kubernetes, EKS, Docker, Helm, Kustomize, ArgoCD, Argo Workflows, OpenShift, Rancher, Istio, NGINX Ingress, Traefik, Cert-Manager, External-DNS, External Secrets, Sealed Secrets, Lens, k9s, RBAC, HPA — blue-green and canary deploys, GitOps. Familiar with AKS, GKE.
Terraform, Terragrunt, Terraform Cloud / Enterprise, Atlantis, Spacelift, Pulumi, AWS CDK, CloudFormation, Ansible, Puppet, Chef — policy-driven change workflows.
Jenkins, GitLab CI / CD, GitHub Actions, AWS CodePipeline, Azure DevOps, Harness, Argo Workflows, JFrog Artifactory, Nexus, Maven, Gradle, Git, Bitbucket — artifact promotion, deployment gates, rollback readiness.
Datadog, PagerDuty, Prometheus, Grafana, Loki, CloudWatch, ELK, Splunk, Dynatrace, New Relic, OpenTelemetry, Fluentd, Fluent Bit, Vector, Logstash, Beats, Statuspage — APM, traces, RUM, synthetics, SLI / SLO, MTTR, post-mortems.
Aurora PostgreSQL, Aurora Serverless v2, Aurora DSQL, PostgreSQL, MySQL, MariaDB, DynamoDB, MongoDB, Cassandra, Redis, ElastiCache, etcd, Memcached, OpenSearch, Solr, Hazelcast, ActiveMQ, Kafka, MSK, RabbitMQ, Athena, Glue, Kinesis, Firehose — query tuning, EXPLAIN plans, index audits, parameter-group hardening, blue-green engine upgrades.
IAM least-privilege, KMS, HashiCorp Vault, Azure Key Vault, SonarQube, Fortify, SAST, DAST — secrets management, audit logging, log4j2 hardening, OpenSSF Scorecard.
VPC, ALB / NLB, Route 53, CloudFront, CloudFront Functions, Lambda@Edge, Cloudflare, Akamai, OpenVPN, Site-to-Site VPN, AWS Direct Connect, Transit Gateway, PrivateLink — security groups, NACLs, ALB session stickiness, ALB IP rotation.
PagerDuty, ServiceNow, Jira, Confluence, Salesforce, Slack workflows, Microsoft Teams — incident response, change management, customer escalation, runbook authoring.
Python (boto3, python-docx), Bash / Zsh, SQL, YAML, JSON, HCL, Groovy, jq, yq, Java (read-level, bytecode / javap forensics).
Azure (AKS, Azure DevOps, Key Vault, Container Instances) and GCP (GKE, Compute Engine, Cloud Storage, Cloud IAM) for cross-cloud operations.
Cost Optimization (FinOps), Performance Engineering (JMeter, k6), Postgres / MySQL tuning, JVM bytecode forensics, capacity planning, disaster recovery, change management, customer escalation communication.
04 — Experience
Drag — same incident class, before vs. after the parallel diagnostic harness shipped (Q3 2024).
Hyland · Remote / Bengaluru, India
xargs -P 8 that fans 9 health queries across 14 environments at once. Triage time: ~80 minutes → 1 minute 34 seconds. Same-day root-cause work on platform-wide performance issues became possible.GO / WAIT / STOP verdict before any production swap.Pirimid Fintech (regulated banking technology) · Bengaluru, India / Remote
TivonaGlobal Technologies (managed cloud services) · Chennai, India
Early experience: Software Engineering internships at XtroviX Technologies (Pune, Nov 2019 – May 2020) and SmartDigital WebNMobile (Nagpur, May – Jun 2018) — Java, Python, C++, Git, Linux, APIs, SDLC workflows.
05 — Projects
Designed AWS and offline banking infrastructure using ECS, ECR, CloudFront, EC2, VPC, ALB, RDS, Redis, S3, IAM, Route 53, Terraform / Ansible automation, Nginx, HAProxy, Patroni, PostgreSQL, Prometheus, Grafana, Loki, and ETCD.
Built containerized Java deployment using Azure Container Instances, Application Gateway, PostgreSQL, GitLab CI / CD, and equivalent AWS ECS / VPC / ALB / ECR / RDS patterns.
Migrated a live chat-room application from on-premises infrastructure to AWS using EC2, Lambda, ALB, Auto Scaling, VPC networking, RDS PostgreSQL read replicas, Nginx, disaster recovery, and GitHub Actions CI / CD.
Implemented GitLab CI / CD and Terraform-driven deployment for Java services using Maven builds, Selenium tests, Docker image publishing to ECR, ECS task definitions, CloudFront delivery, S3 static hosting, Redis, RDS, and ALB routing.
06 — Education
Malayalamnative Englishfluent Hindifluent Marathiunderstand
06 — Certifications
07 — Contact
Open to senior SRE, DevOps, or Platform Engineering roles. Happy to relocate. LinkedIn is the fastest way to reach me; my email is in the resume PDF.
StringIndexOutOfBoundsException
// javap -p -c -v org.springframework.transaction.support.TransactionSynchronizationUtils at org.springframework.transaction.support.TransactionSynchronizationUtils .triggerBeforeCommit(TransactionSynchronizationUtils.java:81) at org.alfresco.repo.transaction.RetryingTransactionHelper .doInTransaction(RetryingTransactionHelper.java:489) java.lang.StringIndexOutOfBoundsException: begin 0, end -1, length 7 at java.base/java.lang.String.checkBoundsBeginEnd(String.java:4604) at java.base/java.lang.String.substring(String.java:2706) at com.customer.audit.AuditNodeRefHasher.hash(AuditNodeRefHasher.java:42) // Retry loop swallows the exception, rolls back, then commits the txn anyway. // Net effect: 147,513 duplicate documents over 18h before alarm fired. // Fix proposal: bound-check substring(0, refLen-1) before invocation, // fail fast on retry, do NOT swallow.