Pascal Masila

I've evaluated AI-generated responses at AfterQuery using detailed rubrics, scoring outputs and documenting clear justification for every decision. I also contributed to benchmark-task workflows that guide and assess a model’s path toward an intended answer.

At AtlasCapture, Mindrift, and Remotasks, I've annotated video frames and images, identified actions and objects, evaluated image-generation prompt adherence, and applied bounding-box and cuboid annotations. I focus on visually similar edge cases and consistent guideline application because annotation accuracy directly affects downstream model use.

I've also tested web and mobile applications with Test IO, reproducing bugs, documenting evidence, and reporting issues to required standards. My technical foundation includes Python, SQL, Pandas, NumPy, JSON, Git, React, Tailwind CSS, REST APIs, and ongoing CVAT learning.

In IT support roles at MGM Muthu Hotels and Jomo Kenyatta International Airport, I supported hardware, software, network connectivity, access controls, and infrastructure monitoring. I bring detailed documentation, disciplined remote work habits, and careful execution under structured instructions.