AWS ECR

• Deployed a database migration from RDS PostgreSQL to Aurora Global PostgreSQL using AWS DMS resulting in a 3x increase in throughput and a seamless transition of data with minimal downtime.

• Developed and productionized a dynamic QR code generator REST API in GoLang, incorporating feature-oriented programming; images are securely loaded to Amazon S3 & the component is deployed on AWS EC2, resulting in annual savings of 2 million sheets of paper and $700,000 in printing costs.

• Built a component to detect multiple or duplicate letters, alerting on-call engineers of potential production failures using AWS CloudWatch, Java, and AWS ECS, backed by MongoDB.

• Spearheaded the design, development, testing, and deployment of a serverless Go API powered by AWS to retrieve current credit line increase data resulting in a 99% cost reduction compared to the prior API.

• Developed Python web scrapers to extract data on prospective players from deprecated vendors.

• Utilized Google Cloud services to store and manage scraped data in secure and accessible buckets.

• Automated scraping process using GET and POST requests in Python with Apache Airflow to optimize efficiency and streamline data aggregation and integration into SQL-based databases with SQLAlchemy.

• Devised image processing algorithms to convert thousands of ground truth low-light images of planted seeds into high-quality images utilized for labeling training data via LabelMe.

• Constructed a state-of-the-art convolutional neural network using YOLOv4 to identify seed planting and predict crop yield based on critical factors such as seed orientation, trench depth, and soil composition.

• Created an end-to-end server infrastructure for seamless transmission of Deere's real-time sensor data.