Devops

Led and managed the six-member Infra platform team, which comprised Kubernetes, network, infra tooling, and shared services, which included senior and principal engineers. Led the Tiket north-south architecture modernisation project, which introduced Kubernetes envoy gateway, a custom extension service, and a plugin written in Golang. This project reduced hops from 7 to 4. Restructured the ELK stack, Conducted Java runtime performance POCs using GraalVM, EPP, and Azul Prime Java Runtimes, and Implemented Redis Sentinel cluster.

10% latency reduction by implementing an nginx caching layer ahead of the Ruby on Rails monolith service.

Created a real-time multiplayer game, https://rummymatrix.com, starting from the ground up. Developed both frontend and backend components using node.js and crafty.js for the game engine.

Achieved a remarkable 10x acceleration in environment provisioning and deployment by introducing CI/CD pipelines through Jenkins and Python. Created a comprehensive Continuity Engine comprising an Infra, Integration, and Deploy Engine. Executed a company-wide security enhancement strategy, optimising AWS cloud hosting with industry best practices. Implemented a fault-tolerant Secret Management solution using HashiCorp Vault.

Achieved a 30% reduction in Mean Time to Recovery (MTTR) by creating a Python-based web application using Django and MySQL. This innovative tool streamlined operational tasks crucial for incident management. Spearheaded a 4-member Site Reliability Engineering (SRE) team, successfully overseeing critical projects including TLS 1.0 deprecation, seamless migration of secrets to Vault, and implementing auto-remediation via SaltStack.

Enhanced ELK stack reliability and scalability by redesigning the data pipeline with Kafka, resulting in a remarkable 200% reduction in ELK cluster downtime. Streamlined engineer onboarding with an impressive 400% time reduction by creating an automated User management system using a Python-based web control panel powered by Django, Redis, and MySQL.

Optimised the Scala-based bloom filter service, significantly enhanced search result latency by over 80%. This optimisation involved swap as a supplementary store for the in-memory data structure. Achieved seamless deployments with zero downtime by enhancing the Scala-written Katta (Indexing) service by integrating a Pluggable Software availability detection mechanism.

Reduced metrics system incidents tenfold by migrating metrics collection from Graphite to OpenTSDB.

Drove an 80% reduction in completion latency for customer support backup requests by fully automating a new backup process using Perl.

Implemented infrastructure configuration management using Puppet and Perl-based tools, resulting in a reduced configuration error rate of 20%. Reduced custom graph creation latency tenfold by developing a PHP web interface for Ganglia featuring an innovative web-based command-line interface.

Optimised server bootstrapping efficiency by 30% by rectifying Perl-based automation related to the PXE-based bootstrapping process. Part of the team managing around 8k Linux servers, handled highly critical on-call rotation.