Ugboma E.

Key Skills

Software

Mindrift

Top Subject Matter

Ego-centric interactions and object-related video events

Multimodal AI training data (visual-grounded prompts, evaluations, and image transformation instructions)

Top Data Types

Video

Text

Image

Top Task Types

Tracking

Prompt + Response Writing (SFT)

Freelancer Overview

Video Data Annotator at Atlas Capture. Core strengths include N, A, and Mindrift. Education includes Bachelor of Engineering, University of Ibadan. AI-training focus includes data types such as Video and Text and labeling workflows including Tracking and Prompt + Response Writing (SFT).

Labeling Experience

AI Data Annotator at Mindrift

MindriftTextPrompt + Response Writing (SFT)

Created structured prompts and corresponding responses aligned with visual inputs and task guidelines for AI training or evaluation workflows. Evaluated AI outputs to verify accuracy, relevance, and consistency with the given instructions. Performed image transformation tasks, including modifying patterned objects and writing precise prompts to restore visual patterns while avoiding assumptions. • Authored guideline-aligned prompt-response pairs • Verified AI output quality via accuracy/relevance/consistency checks • Completed image transformation and pattern restoration prompt tasks • Followed task instructions strictly in a remote setting

Not specified

Video Data Annotator at Atlas Capture

VideoTracking

Conducted fine-grained ego-centric video annotation by defining precise start and end timestamps for interactions between an individual and objects. Labeled multimedia content to support machine learning model training requirements while maintaining consistent outputs per provided guidelines. Ensured high annotation accuracy in time-sensitive workflows and delivered results reliably while working independently. • Precise start/end timestamping for ego-centric interactions • Followed structured annotation guidelines for consistency • Performed frame-level review to maintain label accuracy • Maintained quality across independent, time-sensitive tasks

Not specified

Education