Mina Naguib Zekry G.

Key Skills

Software

Other

Micro1

Mercor

Top Subject Matter

Creating and evaluating tool-use trajectories and multi-step function-calling workflows for AI agents.

Creating, reviewing, and grading Python/data science tasks using clear rubrics.

Creating bounding boxes and evaluating visual annotations against detailed rubrics.

Top Data Types

Computer Code Programming

Image

Top Task Types

Prompt + Response Writing (SFT)

Function Calling

Bounding Box

RLHF

Computer Programming/Coding

Freelancer Overview

I have over 3.5 years of total experience in data labeling and AI training data, including 2.5 years with Turing, where I worked on both SFT and RLHF projects focused on teaching models to write, reason through, and execute Python code for data science tasks. I also worked on agent training by creating tool-use trajectories that simulate realistic sequences of function calls, helping models learn how to interact with tools effectively. Beyond Turing, I worked with XLAIRS ServicesAI across multiple AI training initiatives, including serving as a TPM for a project focused on creating function-calling trajectories for agent training. I also contributed to projects involving data science task creation and grading rubric design. At Mercor, I began as a trainer on data science tasks, advanced to reviewer, and eventually became an EPM (Experienced Project Manager). I am currently contributing to Micro1 as well, continuing to support AI training and evaluation work.

Labeling Experience

Data Science Quality Assurance and Calibration Specialist

Computer Code ProgrammingComputer Programming/Coding

I contributed to a two projects centered on training agents to execute data science workflows in Jupyter Notebooks. My responsibilities included: • Reviewed completed tasks to ensure quality and readiness before delivery. • Flagged ambiguities early and helped resolve them for consistency. • Writing comprehensive evaluation rubrics covering all user requirements along with the expected success criteria. • Designed rubrics to be fair, robust, and generalizable across multiple valid solution paths without overfitting to a single method or implementation style.

2026 - 2026

Technical Project Manager (TPM)

Computer Code ProgrammingFunction Calling

I served as Technical Project Manager for a short-term project training models to execute agentic workflows using multi-step function-calling trajectories. The role focused on defining task instructions and examples so annotators could generate high-quality training tasks with correct reasoning per step. I also reviewed completed tasks to ensure they met the project’s quality standards. • Acted as primary point of contact for the client. • Developed clear project instructions and guidelines for annotators. • Created example tasks and highlighted common pitfalls for task authoring. • Reviewed annotators’ completed tasks against quality standards.

2025 - 2025

Data Science Experienced Project Manager (EPM)

Computer Code ProgrammingPrompt + Response Writing (SFT)

I participated in a one-month sprint training agents to create and debug Jupyter notebooks across data analysis, machine learning, deep learning, and scientific computing. I began as an annotator, then transitioned to reviewer, and later to project manager, with responsibilities evolving across phases. The work included producing supervised fine-tuning tasks and maintaining project guidance to keep contributors aligned on quality. • Created supervised fine-tuning (SFT) tasks as an annotator. • Reviewed task quality and provided guidance as a reviewer. • Created walkthrough recordings and maintained project instructions as an EPM. • Led QA sessions to clarify requirements and unblock contributors.

2025 - 2025

Python Developer for Data Science

OtherComputer Code ProgrammingPrompt + Response Writing (SFT)

• Developed high-quality training data for a leading AI company through Supervised Fine-Tuning (SFT). • Provided human feedback for model optimization via Reinforcement Learning from Human Feedback (RLHF). • While most contributions focused on data science tasks, I also created tasks that trained LLMs and agents on general problem-solving using Python, reasoning, function-calling, and online search workflows.

2022 - 2023

Education

Work History