Sarabjot Singh

Dynamic DevOps Engineer with a proven track record at Jungleworks, specializing in AWS cloud architecture and infrastructure automation. Expert in Terraform and CI/CD pipelines, I excel in optimizing high-availability systems while fostering collaboration across teams. Passionate about leveraging innovative solutions to enhance performance and reliability in SaaS environments.

Automation & Infrastructure Optimization, Photography & Visual Documentation, Technical Blogging & Knowledge Sharing

Tookan (Delivery Tracking Platform):, - Frontend hosted via Amazon S3 with CloudFront distribution, using Angular served by HTTPD web server., - Backend microservices deployed in private subnets with Auto Scaling groups; two EC2 instances outside autoscaling for manual management, rest using spot instances., - Load balancing handled by HAProxy, dynamically updated through a Lambda function triggered by API Gateway. The Lambda uses Boto3 to add/remove private IPs into HAProxy templates, ensuring seamless service discovery., - Custom shell scripts monitor load average, triggering CloudWatch alarms and autoscaling actions., - SSL/TLS v1.3 encryption managed using Certbot with an automated script that provisions wildcard certificates on client onboarding, renews them, and places them securely., - Master-slave database architecture implemented for MySQL and MongoDB, ensuring data replication, high availability, and performance., - Redis used for fast-access data; DynamoDB used for temporary fleet/location data., Yelo & Hippo (SaaS Platforms):, - Built with the same core architecture as Tookan, except using NGINX + HAProxy for web server and load balancing., - Continuous delivery implemented using Jenkins/GitLab CI, Dockerized deployments across autoscaled EC2 environments., - Database backends using master-slave setup, with PostgreSQL included for dev/test environments., - SSL lifecycle handled with automation, security via KMS encryption, and monitoring unified across all services., Jugnoo (Uber-like Ride Hailing App):, - Backend services deployed partially on Amazon EKS for service isolation and container orchestration., - Frontend services hosted on NGINX + Apache stack., - Integrated the same HAProxy-based dynamic configuration using Lambda for backend microservice load balancing., - Redis and DynamoDB used similarly as in Tookan; secure DB access enforced with private subnets and KMS., Fugu (Chatbot Platform):, - Real-time messaging infrastructure deployed on HAProxy + NGINX backend., - Focus on scalability and resilience using autoscaling EC2s., - Full monitoring and alerting integrated using ELK + Grafana, with performance metrics and chat throughput dashboards.

• ELK Stack (Elasticsearch,Filebeat, Kibana): Implemented on all servers for centralized logging and deep request tracing.
• Grafana Dashboards: Used across environments for system resource monitoring, container health, and custom metrics.
• Graphite + Collectd: Deployed for infrastructure telemetry and integrated into Grafana for alert visualization.
• Nagios: Used for legacy resource monitoring and incident notifications, enhanced with Twilio call alerting for critical triggers.
• Backup Monitoring: Custom shell scripts validate database and file system backups with failure alerts sent via webhook to Google Chat and email for DevOps team.