Certified AWS Solutions Architect - Associate
Solutions Architect in training, driven by a lifelong fascination with AI and emerging technology. Actively pursuing opportunities to design cloud-native infrastructures that scale, recover, and evolve—what I call VIBEs: Virtual Infrastructures Built for Evolution.
I’m an AWS Certified Solutions Architect with a fullstack development background and hands-on AI/ML expertise. My focus is on designing Virtual Infrastructures Built for Evolution (VIBEs) — cloud-native systems that scale, recover, and continuously adapt to modern business needs.
Through projects ranging from serverless architectures to AI-powered platforms, I’ve built a track record of turning complex challenges into secure, cost-efficient, and intelligent solutions. My passion lies at the intersection of cloud engineering and AI innovation, where I create systems that don’t just work for today, but evolve to meet the demands of tomorrow.
Migrated a local Node.js/SQLite resume app to AWS via Terraform, configuring a private S3 static site, CloudFront with OAI, Route 53 DNS, and ACM certificate. Developed a serverless visitor counter using DynamoDB and Lambda with GitHub Actions CI/CD workflows.
Built a secure, cost-efficient AWS stack with Terraform to host Weaviate and Streamlit platform. Deployed Weaviate via Docker on EC2, integrated OpenAI embeddings for FAQ vectorization, and created a Streamlit frontend for semantic and hybrid search with real-time relevance scoring.
Engineered and trained a video instance segmentation model using YOLOv8, enabling real-time vehicle detection and tracking with TPU acceleration. Executed advanced data augmentation techniques and optimized via ONNX export for enhanced GPU inference in autonomous systems.
Architected and maintained the backend of an AI-powered personal finance app using Node.js and Express in an Agile environment. Embedded OpenAI chatbot for personalized investment advice and leveraged PostgreSQL with Knex for streamlined database management and CI/CD pipelines.
Constructed an object detection pipeline utilizing Facebook's Detectron2 framework to train and evaluate a Faster R-CNN model on custom balloon dataset. Delivered detailed performance insights using COCO's Average Precision metrics and established TensorBoard for training monitoring.
Designed a deep learning pipeline for semantic segmentation using Amazon SageMaker, harnessing AWS cloud services for scalable model training and deployment. Executed data preprocessing, augmentation, and established a real-time inference endpoint for seamless model integration.
Fresh off my AWS Solutions Architect certification, I tackled the Cloud Resume Challenge to build real-world cloud experience. This comprehensive journey covers migrating a legacy site to AWS using Terraform IaC, implementing serverless architecture with Lambda and DynamoDB, setting up CI/CD pipelines, and overcoming DNS, CORS, and IAM challenges along the way.
Enjoyed this deep dive? Follow me on Dev.to for more solutions architecture and AI insights.
Follow on Dev.to