Headline

Senior Software Engineer @ Atlassian | Backend & Platform Engineering | Distributed Systems & Observability | Cloud (AWS/GCP, Kubernetes, Terraform) | Site Reliability & Scalability at Enterprise Scale

About

Senior Software Engineer with 8+ years building large-scale distributed systems, observability and data platforms, and streaming infrastructure. At Atlassian, I own critical pieces of the centralized observability and data platforms scaling Grafana Mimir to 500M+ active time series at 99.99% SLO, cutting $2M/year in infrastructure cost, and operating a 12,000+ job Apache Flink fleet across 12 regions via a Python Kubernetes Operator I designed. More recently, I've been building the AI-native layer on top of this platform: MCP (Model Context Protocol) servers that expose Flink, StarRocks, logs, and runbooks as agent-callable tools, unlocking LLM-driven incident triage, RCA, and guardrailed remediation for on-call teams. I care about the boring fundamentals SLI/SLO design, capacity planning, FinOps, multi-region reliability — and about the frontier where platform engineering meets agentic systems. Currently leading Atlassian's AWS→GCP data platform migration and exploring senior/staff roles where infrastructure scale and AI-assisted operations intersect.

Location

Bengaluru

AI stack

anthropic apis

claude code

claude

cursor

deepseek

glama

Work experience

Senior Software Engineer at Atlassian, Aug 2024 - Present · 2 yrs 1 mo, Bengaluru, Built an MCP (Model Context Protocol) tooling layer exposing Plato's Flink, Paimon, StarRocks, logs, and runbooks as standardized, permissioned, agent-callable tools with RBAC and audit trails enabling AI-assisted incident triage, RCA, and human-approved remediation. Reduced MTTR and eliminated manual context-stitching across on-call rotations. Architected a Kubernetes Operator in Python (kopf) after evaluating sidecar/DaemonSet alternatives; chose the operator pattern for CRD-driven lifecycle handling. Scaled to 12,000+ Flink jobs (1,000+ per region × 12 regions), replacing per-cluster monitoring with unified fleet-wide observability. Deployed and scaled the Flink History Server for post-run diagnostics and SLA tracking across the same footprint. Leading a 4-person team on the AWS → GCP migration of Atlassian's data platform (real-time Flink streaming + batch pipelines, 12 regions), owning cross-cloud auth, GCP Workload Identity Federation (WIF), service account provisioning, data transfer, and end-to-end cluster setup. Owned and scaled Atlassian's centralized Grafana Mimir platform to 500M+ active time series at 15K write QPS / 6K read QPS, supporting 130K+ detectors at 99.99% SLO cutting infra cost by $2M/year and query latency by 40%. Designed a multi-tenant authentication gateway for Mimir ingestion in Go using a custom reverse-proxy architecture, processing 15K write QPS, eliminating cross-tenant exposure risk, and replacing a manual allowlist system. Built a K6 load-testing framework validating capacity against Atlassian-scale SLAs and de-risking cost-optimization rollouts. Pioneered an AI/ML-powered SignalFlow→PromQL translation system at 88% accuracy, eliminating 2,000+ hours of manual migration and accelerating the SignalFx → Prometheus/Grafana transition by 6 months.

Software Engineer III at InMobi Advertising, Jul 2022 - Aug 2024 · 2 yrs 2 mos, Bengaluru, Led 5-person engineering team managing Glance Publishing Platform, improving system throughput from 10K to 15K RPS (50% increase) through strategic code optimization and architectural improvements, directly supporting content delivery to 10M+ daily active users. Engineered distributed job scheduler handling 20K concurrent RSS scraping jobs using Redis, PostgreSQL, and Python on Kubernetes, reducing publisher onboarding time from 3 days to 5 minutes (98% reduction) and increasing content ingestion capacity by 50x. Refactored legacy CMS architecture using Spring Boot microservices and Redis caching, reducing API latency from 850ms to 340ms (60% improvement) and enabling platform to process 200K contents daily with zero infrastructure additions. Architected real-time content delivery pipeline for Glance Spaces using Java and Kafka, processing 50M+ daily events to deliver articles and videos to lock-screen users, contributing to 15% DAU growth and 20% improvement in user satisfaction scores. Conducted comprehensive infrastructure audit using Terraform and ArgoCD, Grafana, identifying and eliminating underutilized resources, reducing monthly cloud costs from $140K to $98K (30% savings) while maintaining performance SLAs.

Software Engineer II at InMobi Advertising, Sep 2021 - Jun 2022 · 10 mos, Bangalore Urban, Karnataka, India, Developed distributed load testing platform using Locust, Python, and Kubernetes, enabling teams to simulate production-level traffic patterns, cutting test setup time from 5 hours to 2 hours (60% reduction) and accelerating release cycles by 35%. Built vendor-agnostic event-driven microservice supporting multiple message brokers (Kafka, PubSub), handling 1000+ subscribers and processing 10M+ daily events, saving development teams 20 engineering hours weekly on integration tasks. Implemented company-wide OpenTelemetry observability stack with Java and Kubernetes, instrumenting 40+ microservices and improving debugging efficiency by 50%, reducing average incident resolution time from 4 hours to 2 hours. Modernized 8 legacy Jersey-based services to Spring Boot, improving developer velocity by 45% through better tooling and reducing build/deployment times from 12 minutes to 9 minutes (25% faster), accelerating feature delivery. Technology: Java, Spring Boot, Docker, Kubernetes, Azure, CI/CD, Kafka, GCP

Senior Software Engineer at Tata 1mg, Oct 2020 - Sep 2021 · 1 yr, Gurugram, Haryana, India, Architected digital health records platform using FastAPI and PostgreSQL, enabling real-time lab report and prescription management for 50K+ monthly patients, improving data accessibility by 80% and supporting Tata 1MG’s telemedicine expansion. Created production-ready FastAPI microservice boilerplate with comprehensive unit tests and CI/CD integration, adopted by 10+ teams and doubling developer productivity by reducing new service setup from 2 days to 4 hours. Built a centralized communication platform using Python and SQS/Kafka, handling 100K+ daily notifications (email, SMS, WhatsApp). Streamlined user notifications for multiple teams, enhancing customer engagement, operational efficiency and delivery reliability from 80% to 99.99%. Technology Used: Python, Postgres, Cassandra, Redis, Docker, Kubernetes, AWS, FastAPI, Sanic, RabbitMQ

Software Engineer at Tata 1mg, Feb 2018 - Sep 2020 · 2 yrs 8 mos, Gurgaon, Haryana, India, Working on designing scalable microservices which help in Automating the process of ML Model training to deployment in production. Reduce manual effort in training ML model by 80%. (Technology Used: Docker, Kubernetes, Postgres, Celery, FastAPI, Python) Developed a recommendation engine for recently viewed items and trending products, and enhanced user experience, resulting in a 30% uplift in conversion rates and a 20% increase in click-through rates. Launched Targeted Automated User Surveys, disseminating healthcare information and generating health assessments, increased user engagement by 45% and awareness through gamified experiences, targeting based on user cohorts and transaction history. Ref. Link Engineered a subscription-based Diabetes care plan system, incorporating DB design, API flow, and microservice, attracted 3500+ new daily users, boosting MVP and customer retention, and became Hero Product of Tata 1mg. Integrated different 3rd party delivery and business partner’s APIs (ShadowFax, Buledart, Expressbees, Ecom-Express, SAP, Salesforce) to improve logistic and financial services. Introduced a real-time Agent Monitoring System to track performance metrics of pharmacists, customer care, and prescription reviewers improved overall productivity by 15% and enabled efficient management oversight

Software Engineer Internship at Tata Steel, Jun 2016 - Jul 2016 · 2 mos, Kharagpur Area, India, Develop a Web Portal For Online Complaints Management and interstate in Android Platform Complaints Management System provides an online way of registering and solving the housing problems faced by the employees of Tata Steel (Bearing Division) Kharagpur. It is an online portal for addressing grievances of the employee for numbers of purpose like- any small electrical or civil issues e.g broken window problem and any problem related to housing maintenance.  Current Scenario In current scenario there is very less system operating for registration of complaints. Presently complaints registration process is done through telephone or employee must go to security main gate for registering their complaints. This process requires a manual effort and can lead to confusion for the employees. So for efficient management of complaints and to reduce the response time management decided to design a web portal and an android app for this purpose.  Objectives  Managing complaints efficiently and reducing response time.  Maintaining the complaints record of each employee. Technology Used: PHP, AJAX, JavaScript, Android, Bootstrap, Oracle 11g, jQuery.

Software Engineer Internship at Tata Steel, May 2016 - Jun 2016 · 2 mos, Kharagpur Area, India, Designed a Shift Roster System for Tata Steel Bearimgs Division Shift roaster is very important for any Industry. It will keep all the record of employee‘s shifts. This shift roaster system is developed for Tata Steel (Bearings Division). From a long observation it was noted that employees are not present in their assigned shift which results in loss in production. So administration decided to develop such a system which can help in following:  The system will help in calculating expected manpower and available manpower in any particular day according to shift and department.  The system will give information absenteeism according to shift as well as department for a month or day.  The System will give the shift details for each employee and also allow authorized person to modify shift details. This System help administration as well as organization in number of ways:  The System will give information about absenteeism it will help administration to take necessary action to reduce the absenteeism.  The System will give information about man power it will help in increasing as well as maintaining the production.  Via cutting the reliance on expensive, short notice workers to fill gaps in schedules when it may appear to be the only solution. Technology Used:  RDBMS: Oracle 10g  Programing language: PL/SQL  Oracle Developer

Education

Bachelor’s Degree, B.Tech Information Technology, National Institute of Technology Durgapur

Diploma In Engineering, Electronics And Instrumentation, Harasankar Bhattacharya Institute Of Technology And Mining Ranigung