Abdulganiu A.

Key Skills

Software

Mindrift

Top Subject Matter

Home/household activities

computer vision

object recognition

Top Data Types

Video

Audio

Image

Top Task Types

Action Recognition

Transcription

Audio Recording

Bounding Box

Object Detection

Data Collection

Freelancer Overview

AI Data Contributor – Video Annotation & Data Collection | Mindrift. Brings 6+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Mindrift, Uber AI, and Luel. Education includes Bachelor of Technology in Statistics, Federal University of Technology, Minna (2024). AI-training focus includes data types such as Video and Audio and labeling workflows including Action Recognition, Transcription, and Audio Recording.

Labeling Experience

Video Annotation Specialist (Egocentric Tasks) | Atlas

VideoAction Recognition

Annotated first-person perspective (egocentric) videos showing humans completing physical tasks for action recognition AI models. Reviewed text annotations segment-by-segment to identify and correct mislabeled data and maintain dataset accuracy. Applied domain knowledge to ensure labeling consistency across large video datasets. • Labeled egocentric task videos for action recognition • Performed segment-by-segment annotation review and corrections • Evaluated annotation consistency using domain understanding • Maintained labeling standards for large-scale video datasets

2026 - Present

Screen Recording & Conversational Audio Contributor | Luel

AudioAudio Recording

Captured conversational audio inputs and supported structured partner-recording sessions to generate natural dialogue datasets for speech AI models. Contributed to recording and dataset creation workflows by providing real-world conversational audio examples. Helped ensure the resulting audio data supports effective speech model training. • Recorded conversational audio as part of partner sessions • Generated natural dialogue datasets for speech AI • Captured structured audio data for training pipelines • Supported collection quality for speech-model labeling

2026 - Present

Audio Transcription Contributor | Uber AI

AudioTranscription

Transcribed audio recordings with high accuracy to generate labeled speech datasets for AI voice and NLP training. Maintained transcription quality standards to ensure consistent and correct outputs across diverse audio samples. Helped produce reliable labeled text for downstream speech and language models. • Transcribed audio recordings for labeled speech datasets • Ensured consistency and correctness across varying audio samples • Applied quality standards during transcription work • Produced training-ready text labels for AI/NLP pipelines

2026 - Present

AI Data Contributor – Video Annotation & Data Collection | Mindrift

MindriftVideoAction Recognition

Provided first-person (egocentric) video recordings of household tasks to help train AI models on hand movements and home activities. Uploaded and curated annotated photo datasets and performed content/title preference labeling to improve computer vision understanding and AI content ranking algorithms. Contributed to dataset preparation and annotation quality improvement for machine learning pipelines. • Recorded egocentric household-task videos for model training • Curated annotated photo datasets for computer vision/object recognition • Completed title preference and content labeling for ranking improvements • Supported data collection and annotation review workflows

2026 - Present

Education

Work History