• Used CDK to deploy an EKS cluster for computer vision classification application used to assess quality of chemical drugs, and worked closely with the developers to deploy Prometheus for business metrics as well as cluster monitoring. • Contributed to a turnkey Kubernetes deployment tool highlighted on the awesome Talos repository on GitHub.• Leveraged AWS CDK to automate the deployments of all the workloads of internal software projects. • Implemented a virtualization hypervisor cluster based on Proxmox.• Implemented the Terraform codebase for VM and networking configuration on Proxmox virtual environment consuming the Proxmox provider.• Implemented GitOps Kubernetes deployment codebase for deployment of Prometheus, Netris and ArgoCD operators consuming corresponding providers as well as GitHub providers.• Implemented a sample service with Kubernetes manifests as well as ArgoCD application manifests to automate the continuous delivery with ArgoCD.• Designed and Implemented the complex CI workflow to adhere GitOps principles for Kubernetes cluster management.

• Implemented complex CI pipelines for deployment of PR environments.• Implemented GitOps deployment workflows with complex versioning and branching strategies.• Deployed hybrid infrastructure using On-Prem machines and AWS Direct Connect.• Maintained the entire Terraform Codebase.• Designed microservices architecture using FastAPI, andFlask for REST APIs, ensuring performance and maintainability.• Utilized PostgreSQL, and MySQL for database managementusing SQLAlchemy, ensuring efficient and secure data handling.• Employed dependency injection promoting modularity andcode reusability.• Developed a data provider accepting SOLID principles to work(upload, download, crop, etc.) images in various sources like AWS,DeepLake, HuggingFace, and GCP.• Created a Python package for the project in PyPI.• Created and managed EC2 instances, implementing AWS SESfor email services.• Dockerized the entire application and deployed it on AWS.• Deployed AI-based APIs on GPU-enabled EC2 instances,facilitating high-performance computation for machine learningtasks.• Implemented a robust system where the main applicationinteracts with EC2 instances to run tasks, download results, andmanage instance lifecycle via multiprocessing.• Designed and implemented multithreading for efficient imagedownloading, enhancing the performance of ML model training.• Implemented JWT-based authentication, ensuring secure accesscontrol across services.• Integrated Stripe for payment processing, enhancing the tool'smonetization capabilities.

• A Telegram bot for newsfeed collection.• Scraped real estate pricing information from various websites and trained a recommendation model for my website.• A 3D Model marketplace website.

• Maintained the network equipment and computers of the Vayk’s school campus.• Managed Active Directory of Vayk’s school.• Developed Access programs for school grade reporting.