Resume

Project: Autonomous Infrastructure AI Agent (System Architecture Phase)

Architecting a production-grade, AI-driven deployment engine designed to translate natural language into secure, cost-optimized Infrastructure-as-Code (IaC).

• System Architecture & Logic Design: Designed the core operating model for an autonomous deployment agent prioritizing opinionated, industry-standard Terraform module generation over raw resource blocks to ensure scalable, zero-trust cloud architectures.
• Safety & Execution Guardrails: Conceptualized a rigorous pre-deployment validation pipeline requiring mandatory terraform plan dry-runs, human-readable change summaries, and explicit user confirmation to prevent accidental infrastructure drift.
• Destructive Action Interception: Architected a custom interception layer designed to parse execution plans and strictly block operations violating project boundaries or attempting irreversible destructive actions.
• Technologies and concepts: Terraform (HCL), AWS, Kubernetes, Linux, Docker.

Project: Agenthic — AI Agent Infrastructure Platform

Architected a purpose-built infrastructure platform for AI agent lifecycle management. Designed a deterministic knowledge retrieval system using custom .index and .map formats, replacing probabilistic RAG pipelines to enable precise document retrieval and live infrastructure dependency tracking.

• Deterministic Retrieval Architecture: Designed a hierarchical indexing engine to replace traditional RAG pipelines, mapping table of contents headings to exact file references without relying on vector embeddings.
• Infrastructure State Memory: Architected a live dependency tracking layer by parsing Terraform state files into dependency graphs, defining relationship mappings (REQUIRES, DEPENDS_ON) to ensure lifecycle consistency.
• Safety & Execution Guardrails: Conceptualized a custom Go-based Terraform interception system designed to strictly block autonomous code deletion risks and enforce auto-remediation safety nets prior to execution.
• Technologies and concepts: Terraform, AWS, Kubernetes, GitHub Actions, CI/CD pipelines.

Engineered production-grade cloud infrastructure and CI/CD workflows during a dedicated technical upskilling period, prioritizing modularity, zero-downtime migrations, and automated deployments.

Infrastructure as Code & CI/CD

• Standardized infrastructure delivery by creating reusable Terraform modules for multi-tier architectures.
• Automated the deployment of a dynamic Node.js web application by engineering a CI/CD pipeline integrating Jenkins, Docker, and Docker Compose.
• Architected a highly available Amazon EKS cluster for AI microservices, deploying the AWS Load Balancer Controller via native Linux commands and kubectl manifests to maintain granular infrastructure control.

Cloud Migrations & Data Architecture

• Executed a cross-cloud migration of MySQL database from Azure to AWS RDS using AWS DMS (Full Load + CDC), executing end-to-end transfers with zero data loss.

Serverless Workflows & Edge Delivery

• Developed serverless workflows utilizing AWS Lambda, triggered by S3 events, to automate file compression and dispatch email notifications upon object creation.
• Deployed a highly available static website on Amazon S3, routing traffic via Route 53 and optimizing edge performance utilizing CloudFront caching.

• Automated infrastructure provisioning for an e-commerce platform using Terraform (EC2, RDS, VPC).
• Designed a multi-tier fault-tolerant architecture on AWS, ensuring high availability.
• Deployed a Node.js web application using Elastic Beanstalk, Auto Scaling, and CloudWatch Alarms.
• Automated CI/CD pipelines for multiple clients using AWS CodePipeline and GitHub.
• Migrated 2+ on-premises MySQL database to AWS Aurora via AWS DMS.
• Migrated DNS from GoDaddy to AWS Route 53.
• Optimized AWS cost management by identifying and terminating idle resources (EC2, RDS, ALB, Elastic IPs).
• Configured CloudFront caching to reduce data transfer costs and improve performance.
• Implemented S3 lifecycle policies to automatically transition data to cost-effective storage classes.