Logistics network visualization
cosmos.com

Revolutionizing Logistics Tech at Delhivery

Working at Delhivery, one of India’s logistics giants, meant solving problems at scale—millions of packages, tight delivery windows, and high expectations. I contributed to systems that didn’t just support operations—they moved the country.


Mission-Critical Platforms

Polymapper: Smart Geolocation for Last-Mile

  • Designed a “nearest facility predictor” using Django Rest Framework and Next.js
  • Integrated Leaflet.js to visualise serviceable zones with interactive maps
  • The goal was to cut misrouted deliveries—and we did

Flare: Automation in Action

  • Engineered an intelligent dispatch system using React and Flask
  • Introduced RabbitMQ and Celery to decouple processing from the UI
  • The system could assign packages in real-time based on capacity and geography

Reimagining Legacy & Infrastructure

Reactifying the Stack

  • Spearheaded the migration from Angular to React
  • Introduced Tanstack Query to simplify and speed up data fetching
  • The revamped frontend was cleaner, faster, and easier to maintain

Scalable, Cost-Efficient Infrastructure

  • Transitioned infrastructure from AWS to GCP, orchestrated via Kubernetes
  • Tuned job queues using Celery, improving response times under peak loads
  • Actively collaborated in setting up resilient CI/CD with GitHub Actions, Jenkins, and Devtron

Collaboration and Culture

From mentoring junior developers to resolving critical production incidents, I wore many hats. Pair programming and team knowledge-sharing were part of my weekly rhythm, not a quarterly checkbox.


Engineering Toolchain

graph TD
Infra[Infrastructure] --> K8s[Kubernetes]
Queues --> Rabbit[Message Queue - RabbitMQ]
Queues --> Celery[Task Runner - Celery]
CI/CD --> GH[GitHub Actions]
CI/CD --> Jenkins
CI/CD --> Devtron
Monitoring --> Logs[Elasticsearch Logging]
Testing --> Unit[Jest]
Testing --> E2E[Cypress]