At DeveloperHub Corporation, I engineered and deployed machine learning microservices with Python, Flask, and REST APIs for client-facing applications. I also developed automated data-cleaning and model-validation pipelines that accelerated iteration cycles by 40%.
At Atomcamp, I built and benchmarked classification and regression models using cross-validation and hyperparameter tuning. I developed NLP pipelines for toxic-text classification and created Power BI dashboards and Streamlit applications to present model outputs.
On my Automated Alzheimer’s Staging via Vision Transformers project, I fine-tuned a ViT-B/16 model for four-stage dementia classification and achieved 99.37% accuracy with a 0.99 F1-score. I quantized the model to TFLite for edge inference in a serverless mobile application framework.
