Ezeh V.

Key Skills

Software

Mindrift

Micro1

Toloka

Telus

Top Subject Matter

AI-generated content evaluation across multiple domains

Legal Services & Contract Review

Regulatory Compliance & Risk Analysis

Top Data Types

Text

Document

Image

Top Task Types

Text Summarization

Text Generation

Question Answering

RLHF

Evaluation/Rating

Bounding Box

Classification

Object Detection

Fine-tuning

Transcription

Prompt + Response Writing (SFT)

Freelancer Overview

AI Evaluator & Data Annotator, Mindrift (January 2026 – April 2026). Brings 3+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Mindrift. Education includes Bachelor of Laws, Enugu State University of Science and Technology (2025) and Bachelor of Laws, Nigerian Law School (2027). AI-training focus includes data types such as Text and labeling workflows including Evaluation and Rating.

Labeling Experience

Data annotator at Turing

ImageText Summarization

I worked on a dataset project focused entirely on image and video annotation. My main responsibilities included drawing precise bounding boxes, tracking specific objects frame-by-frame across video clips, and applying descriptive tags based on the project guidelines. Over the course of this 5-month project, I annotated and processed over 500 separate image and video assets. To maintain high data quality, I carefully reviewed edge cases to ensure accurate box placements and consistently met the team's internal quality assurance standards with an accuracy rate of 96%.

2026 - 2026

AI Evaluator & Data Annotator, Mindrift (January 2026 – April 2026)

MindriftText

Performed AI response evaluations by reviewing, rating, and comparing AI-generated answers using structured guidelines. Flagged factual inaccuracies, logical inconsistencies, and instruction-following deviations across multiple task domains. Selected best outputs among multiple model responses with clear written justifications. • Rated and ranked responses based on detailed evaluation rubrics • Compared multiple model outputs and chose the strongest answer • Identified errors in factual accuracy, logic, and compliance • Adapted to updated instructions and evolving evaluation criteria

2026 - 2026

Education

Work History