John Michael Weaver

Developed an AI Medical Supporting System utilizing computer vision and large language models with PyTorch, Claude, MIMI3, AWS SageMaker, and S3 buckets implementing a retrieval-augmented generation pipeline. Created a real-time data analytics dashboard using Snowflake and Python integrated with Apache Airflow for ETL orchestration, reducing report generation time by 70%. Built an LLM-powered chatbot using Hugging Face and DialogFlow deployed on AWS SageMaker with CircleCI pipelines, improving customer satisfaction by 18% and reducing call center load by 40%.

Developed and trained deep learning models for person re-identification, gender classification, and comparative image quality assessment using PyTorch, achieving high accuracy and precision. Employed code optimization techniques which significantly reduced face detection and feature extraction time. Designed a CUDA-based algorithm to find the most similar person with a 98% reduction in processing time.

Developed a medical imagery detection system performing image segmentation on radiology and pathology images to detect anomalies and grade diseases, also providing medication recommendations using SAM on AWS SageMaker. Introduced a customized AI chatbot for banking CRM management deployed on AWS. Developed real-time rain droplet removal prototypes for vehicle cameras using Spark, Apache Kafka, Python, and PyTorch. Implemented ML applications with Django Rest Framework and deployed on AWS Elastic Container. Enhanced system reliability and decoupling using AWS SQS for background operations and designed algorithms for real-time lane marking detection with OpenCV and PyTorch, deploying websites on AWS with CI/CD pipelines via Jenkins.