Home

Diviyam Pathak

Software engineer interested in systems, backend engineering, cloud, automation, and thoughtful software that works well in practice.

About

I am a software engineer with experience across Kubernetes, automation, backend systems, and low-level operating system work. Over time I have been drawn to projects that combine careful debugging, good abstractions, and real-world reliability, whether that means improving team workflows, shipping internal tools, or understanding what happens beneath the surface.

Experience

Recent work

Jan 2026 - Present

Giant Eagle GCC, Bangalore

Associate Software Engineer

  • Standardized deployment patterns using Helm and GitOps workflows, reducing release failures.
  • Developed ingress, service mesh, secrets management, and observability services used by 100+ engineering teams.
  • Integrated hybrid-cloud networking between on-prem Kubernetes clusters and public cloud environments, enabling seamless workload portability.
  • Supported Black Friday and Cyber Week readiness exercises, ensuring platform scalability for traffic spikes exceeding 4x baseline demand.

May 2025 - Jan 2026

Giant Eagle GCC, Bangalore

Graduate Software Trainee

  • Authored operational runbooks and automation scripts that reduced average incident resolution time by 25%.
  • Developed Terraform modules and deployment templates for Kubernetes-based services, reducing infrastructure provisioning time by 3 hours.
  • Implemented monitoring dashboards using Prometheus and Grafana covering 40+ microservices, reducing incident triage time by 35%.

Apr 2024 - Nov 2024

The NetBSD Foundation, Remote

Google Summer of Code Contributor

  • Debugged C code in NetBSD kernel space to improve system reliability and low-level functionality.
  • Built automated test environments using Bash scripting and virtualization for repeatable systems validation.
  • Developed ATF-based test cases to strengthen operating system test coverage and regression detection.

Projects

Selected builds

LLM-Driven Incident Triage Agent

FastMCP, ServiceNow, Python, Prometheus

  • Built an MCP-compliant, email-triggered incident triage agent exposing ServiceNow incident history and CMDB resources to LLM tools.
  • Automated assignment-group recommendation and priority classification via LLMs with least-privilege ServiceNow API access.
  • Evaluated the system on historical incidents, achieving 35% faster triage and 85% assignment accuracy with end-to-end latency under 2 seconds.

Distributed Task Processing Platform

FastAPI, Celery, RabbitMQ, Redis, PostgreSQL, Kubernetes

  • Developed multi-tenant ingestion pipelines for UPI, IMPS, and NEFT transaction processing with schema normalization.
  • Designed scalable distributed task queues with rate limiting, autoscaling workers, and asynchronous reconciliation workflows.
  • Deployed on Docker and Kubernetes, sustaining 200k+ rows per minute with CI pipelines implemented via GitHub Actions.
View repository

Network Latency Tracker

AWS Lambda, API Gateway, DynamoDB, CloudWatch, Terraform

  • Built a serverless monitoring system for scheduled endpoint latency and availability checks every 5 minutes.
  • Designed reusable Terraform modules for Lambda, DynamoDB, API Gateway, and scheduler components across dev and prod environments.
  • Applied least-privilege IAM and on-demand billing for a secure, cost-aware reliability workflow with historical uptime reporting.

Linux Systems Analysis and Dynamic Instrumentation

DynamoRIO, perf, strace, vmstat, iostat, Valgrind

  • Analyzed Linux systems using instruction tracing, syscall analysis, and performance tooling to surface CPU hotspots and scheduling overhead.
  • Traced kernel-user space interactions to identify blocking syscalls, heavy I/O paths, memory pressure, and queue-depth saturation.
  • Evaluated memory correctness and allocation efficiency to detect leaks, invalid accesses, and heap utilization issues.