developer · platform engineer · systems tinkerer

{ Sharran M }

Lead Software Engineer · Platform Engineering & Developer Experience

Staff+ platform engineer with 7+ years building developer platforms, infra automation, and internal tooling for ~1,000 engineers. I build systems that reduce cognitive load for product teams: CI/CD orchestration, infrastructure automation, migration tooling, and operational guardrails.

+91 7502215354sharran.murali@gmail.com
Bangalore, India
msharran.github.io msharran.github.io/devlog linkedin.com/in/sharran-m github.com/msharran
55% CI/CD latency reduction 18 min → 8 min across ECS, EKS, and Lambda workflows.
1 config on-demand EKS path deployment_platform: eks enabled opt-in Kubernetes deployments.
900+ services migrated to Pulumi Migrated 900+ of 1,400 services in 5 months after the migration had stalled.
40% DevEx satisfaction increase Centralized legacy CI validation into GitHub Checks with zero repo-level config.

Profile

Impact-first systems engineering

Recent work includes fleet-scale Pulumi migration, an EKS platform adopted on demand for persistent/data workloads, CI/CD performance optimization, webhook-driven GitHub Checks, and self-service infrastructure systems that reduce downtime, lead time, and manual coordination across platform teams.

Experience

Selected impact
Lead Software Engineer, Platform Engineering — CRED
Jan 2023–Present · Promoted Apr 2025
  • CI/CD performance: Re-architected the global delivery pipeline with parallel stages, persistent artifact caching, and optimized S3 staging, cutting end-to-end latency 55% (18 min → 8 min) across ECS, EKS, and Lambda workflows.
  • Pulumi migration at fleet scale: Migrated 900+ of 1,400 services to Pulumi in 5 months, reviving a stalled migration with automated CloudFormation→Pulumi code generation, dependency-graph inference, Route53 weight-stepping rollout, and LLM-assisted review steps.
  • EKS platform: Built an opt-in ECS→EKS path through one config change (deployment_platform: eks), supporting stateless services and StatefulSet-backed persistent workloads such as event logging and ingestion services. The platform was widely adopted for persistent workloads by data platform and database teams, and used Pulumi Helm deployments to self-host Cassandra, Karpenter, and Istio Gateway through the same in-house YAML interface teams already used for ECS.
  • CI/CD quality guardrails: Stabilized legacy Bash, Go, and Python CI/CD systems by introducing unit tests across critical execution paths, reducing recurring downtime from 3–4 incidents/month to near zero.
  • GitHub Checks modernization: Centralized legacy CI validation into a webhook-driven GitHub Checks service with zero per-repository developer configuration, improving feedback quality and raising DevEx satisfaction by 40%.
  • CI/CD knowledge assistant: Built and released a Skill + CLI assistant that lets local coding agents answer platform questions and safely edit config files using grounded, source-backed context.
  • Project Caterpillar: Designed a Go-based CLI and orchestration layer with an acyclic dependency model, declarative YAML DSL, and Pulumi Automation API engine for safer platform rollouts.
  • Async orchestration: Built a secure SQS + client-go workqueue layer with custom Git-push triggers, enabling pipeline execution across 1,000+ microservices without exposing Jenkins to the internet.
Senior Software Engineer, Site Reliability — Freshworks
Jun 2022–Dec 2022
  • Async operators: Extended Kubernetes operators with asynchronous reconciliation to sync AWS Secrets Manager → Lambda via a custom CRD, enabling automated secret rotation for serverless workloads.
  • Developer tooling: Built secretctl, a Go CLI that gave developers a kubectl-like workflow for secrets and replaced ticket-driven operations with self-service.
Software Engineer, Cloud Infrastructure — Freshworks
Aug 2021–Jun 2022
  • Self-service VPC pipeline: Authored AWS CDK provisioning that cut infrastructure request turnaround from days to minutes.
  • Teleport at scale: Automated Teleport access-proxy deployment across 180+ AWS accounts via a Go-based Jenkins pipeline.
Earlier Experience
2019–2021
  • Techjays: Led iOS delivery to 5+ App Store releases (100K+ installs); built React on ECS Fargate for real-time analytics.
  • Gofrugal: Core contributor to an Android POS app; focused on performance optimization and tablet-to-mobile UI adaptation.

Leadership

Mentorship and cross-org influence

Skills & Certifications

Depth across platform layers
Languages Go, Python, Ruby · Zig & Rust (exploring)
Platform Kubernetes, EKS, Operators, client-go, Docker
IaC Pulumi, Terraform, AWS CDK
AWS EKS, ECS, Lambda, SQS, Route53, Secrets Manager
Observability Datadog APM, Loki, Prometheus, Grafana
AI-assisted MCP servers, agent workflows, incident and risk analysis
Certifications CKA · AWS SAA

Technical Craft

Selected projects and open source
Selected Projects
  • Redis-in-Zig: High-performance Redis server in Zig; featured as the reference solution by CodingChallengesFYI. PR
  • CodingChallengesFYI: Zig and Rust implementations for strace, reverse proxy, load balancer, and a tiny Docker runtime. repo · containers devlog
  • eBPF egress monitor: Go + eBPF network egress monitor with DNS-aware trace output and a detailed devlog. devlog
  • Protohackers: Network-protocol challenge solutions in Go. repo
OSS Involvements
  • Minikube: Added directory support to minikube cp. merged PR
  • Wasmer Ruby: Patched a Bundler 2.0 compatibility bug. merged PR
  • Ghostty: Early tester; reported a macOS shell-integration selection bug. discussion

Education

Formal training
SASTRA Deemed University, Thanjavur — B.Tech in Electronics & Communications Engineering (2015–19)

References, code samples, and project links available on request.