At Anyone AI, I evaluate AI coding agents and code editing models, creating benchmark datasets and annotations that capture real-world developer workflows and expose model failure modes. I’ve maintained an average score of 4.27/5 across more than 200 model comparison and coding benchmark tasks.
At Cocun Smart, I architected and maintained Python and SQL ETL pipelines that achieved 99% data integrity, and deployed time-series forecasting models for logistics planning. In my published Bioinformatics Journal research, I analyzed protein dynamics with Gaussian Network Models and deep learning, including feature ablation studies on 1D-CNN models with attention.
