Shofodun Tobi

Key Skills

Software

Other

Internal/Proprietary Tooling

iMerit

Mindrift

Remotasks

SuperAnnotate

Top Subject Matter

Computer Vision

Speech Recognition

Large Language Model (LLM) Training

Top Data Types

Audio

Video

Image

Text

Document

Top Task Types

Emotion Recognition

Transcription

Action Recognition

Tracking

Evaluation/Rating

Question Answering

Text Generation

Object Detection

Text Summarization

RLHF

Prompt + Response Writing (SFT)

Segmentation

Freelancer Overview

Data Annotator & Reviewer. Core strengths include iMerit, Mindrift, and Other. Education includes Bachelor of Science, University of Benin. AI-training focus includes data types such as Image, Text, and Video and labeling workflows including Evaluation, Rating, and RLHF.

Labeling Experience

AI Pilot Assistant

MindriftTextRLHF

I collaborate with large language models to refine and validate AI-generated text outputs for quality and real-world usability. I design prompts and conduct structured evaluations to ensure factual reliability and contextual accuracy. I curate datasets aligned with defined taxonomies for robust AI model training. • Achieved over 95% acceptance across reviewed outputs. • Conducted systematic fact-checking to reduce hallucinations and improve model outputs. • Designed prompts to enhance LLM response clarity, relevance, and quality. • Structured and organized AI training datasets as per project guidelines.

I collaborate with large language models to refine and validate AI-generated text outputs for quality and real-world usability. I design prompts and conduct structured evaluations to ensure factual reliability and contextual accuracy. I curate datasets aligned with defined taxonomies for robust AI model training. • Achieved over 95% acceptance across reviewed outputs. • Conducted systematic fact-checking to reduce hallucinations and improve model outputs. • Designed prompts to enhance LLM response clarity, relevance, and quality. • Structured and organized AI training datasets as per project guidelines.

2026 - Present

Data Annotator & Reviewer

ImeritImage

I deliver high-volume annotation and evaluation tasks on image and speech datasets for AI and machine learning projects. I serve as a reviewer in ASR transcription workflows, auditing and correcting machine-generated transcripts to enhance verbatim accuracy. I evaluate and rank image outputs for prompt adherence, ensuring quality and consistency in line with project requirements. • Maintained over 95% quality acceptance rates across all assigned tasks. • Applied detailed guidelines and resolved edge cases to minimize QA rejections. • Balanced high throughput and precision in a fast-paced, production environment. • Reviewed and corrected transcripts to achieve up to 99% accuracy.

I deliver high-volume annotation and evaluation tasks on image and speech datasets for AI and machine learning projects. I serve as a reviewer in ASR transcription workflows, auditing and correcting machine-generated transcripts to enhance verbatim accuracy. I evaluate and rank image outputs for prompt adherence, ensuring quality and consistency in line with project requirements. • Maintained over 95% quality acceptance rates across all assigned tasks. • Applied detailed guidelines and resolved edge cases to minimize QA rejections. • Balanced high throughput and precision in a fast-paced, production environment. • Reviewed and corrected transcripts to achieve up to 99% accuracy.

2026 - Present

Video Annotation & Temporal Event Labeling

Internal Proprietary ToolingVideoAction RecognitionTracking

Participating in AI training projects involving the annotation of complex video data capturing robotic tasks and physical interactions. The work is focusing on accurately describing robot actions, object movements, and task progress across time, supporting AI systems for robotics and embodied intelligence. Responsibilities includes frame-accurate video annotation, temporal event labeling, and object tracking across extended video sequences. Special attention is being given to challenging edge cases such as occlusion, motion blur, partial visibility, and overlapping objects. All annotations are being completed in accordance with strict project guidelines and production quality standards. The role requires consistency, precision, and the ability to interpret task intent and execution quality across dynamic visual scenes

Participating in AI training projects involving the annotation of complex video data capturing robotic tasks and physical interactions. The work is focusing on accurately describing robot actions, object movements, and task progress across time, supporting AI systems for robotics and embodied intelligence. Responsibilities includes frame-accurate video annotation, temporal event labeling, and object tracking across extended video sequences. Special attention is being given to challenging edge cases such as occlusion, motion blur, partial visibility, and overlapping objects. All annotations are being completed in accordance with strict project guidelines and production quality standards. The role requires consistency, precision, and the ability to interpret task intent and execution quality across dynamic visual scenes

2025 - Present

Speech Transcription and Audio Data Labeling for ASR Training

OtherAudioEmotion RecognitionTranscription

Contributed to AI training projects focused on building and validating high-quality speech datasets for automatic speech recognition (ASR) and language models. The work involved transcribing diverse audio recordings and transforming them into clear, well-structured, and accurate text suitable for production-level AI training and evaluation. Responsibilities included segmenting audio, labeling speakers, and ensuring precise alignment between spoken content and text. A key part of the role was reviewing and correcting machine-generated transcripts to improve accuracy, resolve timing issues, and maintain contextual and linguistic correctness across datasets. All tasks were completed under strict transcription, quality assurance, and data privacy guidelines. The project required strong attention to detail, consistent application of annotation standards, and iterative revisions based on quality feedback to ensure the final datasets met required accuracy and reliability benchmarks

Contributed to AI training projects focused on building and validating high-quality speech datasets for automatic speech recognition (ASR) and language models. The work involved transcribing diverse audio recordings and transforming them into clear, well-structured, and accurate text suitable for production-level AI training and evaluation. Responsibilities included segmenting audio, labeling speakers, and ensuring precise alignment between spoken content and text. A key part of the role was reviewing and correcting machine-generated transcripts to improve accuracy, resolve timing issues, and maintain contextual and linguistic correctness across datasets. All tasks were completed under strict transcription, quality assurance, and data privacy guidelines. The project required strong attention to detail, consistent application of annotation standards, and iterative revisions based on quality feedback to ensure the final datasets met required accuracy and reliability benchmarks

2025 - 2025

LLM Training & Multimodal Data Annotation

Internal Proprietary ToolingImageQuestion AnsweringText Generation

Contributed to AI training projects focused on improving large language models through high-quality text and image annotation. The work involved annotating, reviewing, and refining datasets used for model training and evaluation, with strict adherence to evolving project guidelines and quality benchmarks. Responsibilities included labeling complex text and image data, evaluating prompt–response pairs, and ranking or correcting model outputs based on relevance, factual accuracy, and instruction compliance. Particular attention was given to identifying ambiguous cases, edge conditions, and guideline gaps that could affect model performance. All tasks were completed within structured quality assurance workflows, including multi-pass reviews and feedback cycles. The project required analytical thinking, consistency, and careful documentation to ensure datasets met production-level AI training standards.

Contributed to AI training projects focused on improving large language models through high-quality text and image annotation. The work involved annotating, reviewing, and refining datasets used for model training and evaluation, with strict adherence to evolving project guidelines and quality benchmarks. Responsibilities included labeling complex text and image data, evaluating prompt–response pairs, and ranking or correcting model outputs based on relevance, factual accuracy, and instruction compliance. Particular attention was given to identifying ambiguous cases, edge conditions, and guideline gaps that could affect model performance. All tasks were completed within structured quality assurance workflows, including multi-pass reviews and feedback cycles. The project required analytical thinking, consistency, and careful documentation to ensure datasets met production-level AI training standards.

2024 - 2025

Education

Work History