Prathamesh Thakur

- Implemented enterprise observability platform (Prometheus, Loki, Grafana, Jaeger, OpenTelemetry) for 10+ microservices, reducing error detection and RCA time by 90% from manual server log review, transformingincident response from hours to minutes. - Deployed a Firebase healthcare analytics app with a real-time Firestore-to-BigQuery pipeline for 9 collections, reducing admin panel development effort by 30% through direct BigQuery API usage instead of custom endpoints. - Automated setup on 10 Ubuntu servers with Ansible, reducing install time from 8 hours to 30-60 minutes; setup PostgreSQL backups and Slack alerts. - Built CI/CD pipelines on Azure DevOps for 5 Java microservices and frontend with self-hosted agents, reducing deployment time by 30% and integrating SonarQube for code quality gates and test coverage reporting, enabling faster feedback loops and improved code quality across 4 environments. - Improved response times by 30% migrating from Windows IIS to Linux nginx reverse proxy, managing 5 Tomcat instances with SSL/TLS.

- Achieved a 10% reduction in monthly AWS costs by migrating analytics from DynamoDB to self-managed MongoDB on EKS with automated backups, cleaning S3 storage from 450GB to 120MB. - Reduced infrastructure costs by $20K annually (400% ROI) by architecting an on-premises Kubernetes cluster with 3 Linux nodes instead of cloud hosting, while maintaining 99.9% uptime and supporting 15+ microservices in non-production environments. - Administered production EKS cluster (5 nodes, 15+ microservices) with complete AWS infrastructure management, including VPC, IAM, security groups, CloudWatch, and S3, maintaining high availability for production applications. - Self-hosted critical infrastructure (GitLab, Docmost, ArgoCD) with S3 backups and custom registries, eliminating SaaS subscription costs and third-party dependencies. - Implemented GitLab CI/CD pipelines for 15+ microservice repositories, shrinking deployment time from hours to 5 minutes. - Created CloudFormation templates for S3, CloudFront, Route53, and ACM, reducing setup from hours to 30 minutes across four environments. - Designed monitoring dashboards with Slack alerts for HTTP errors and adopted Istio, Sealed Secrets, and FluxCD for better security and GitOps.

- Architected and developed an OpenAI-integrated microservice for automated job description and professional email generation with strict system prompt controls and anti-misuse safeguards, delivered as a production-ready paid feature with complete API integration. - Designed PostgreSQL database schema capturing request/response history, token usage tracking, and a user feedback mechanism enabling output regeneration with context awareness, optimising for query performance and audit compliance.