Open to opportunities

Daniel Adeshina

@danieladeshina1

Message

AI Data Annotator and RLHF evaluator specializing in truthfulness, safety, and large-scale dataset labeling.

Nigeria

Message

What I'm looking for

I’m looking for a role where I can evaluate LLM outputs for truthfulness, safety, and instruction-following, maintain high guideline compliance, and contribute to high-quality dataset labeling with fast-paced, collaborative teams.

I’m an AI Data Annotator & Content Evaluator focused on RLHF evaluation and prompt-response quality. I rate and refine Large Language Model (LLM) outputs for truthfulness, safety, instruction-following, and conciseness, flagging factual errors, reasoning fallacies, and safety violations with written justifications.

I’ve processed 1,000+ AI-generated outputs while maintaining “98%+ accuracy,” strictly following continuously evolving client annotation guidelines. Previously, as a Junior Data & QA Specialist, I performed data quality checks for missing values, duplications, and formatting errors, and standardized records with Excel and basic Python scripts to improve data entry processing speeds by 25%.

Experience

Work history, roles, and key accomplishments

Current

AI Data Annotator & Content Evaluator

Current

Nexus Data Solutions

Jan 2026 - Present (7 months)

Evaluated and rated 1,000+ AI-generated text outputs for truthfulness, safety, instruction-following, and conciseness for LLMs. Flagged factual errors and safety violations with written justifications while maintaining 98%+ labeling accuracy.

Sentiment Analysis Data Annotation Fact Checking Hallucination Detection Python SQL Labelbox CVAT Markdown Git Github AI Data Annotation Bounding Box Labeling Quality Assurance Excel Google Sheets Time Management Critical Thinking Data Quality Assurance

Junior Data & QA Specialist

Apex Technologies Ltd

Jun 2024 - Dec 2025 (1 year 6 months)

Performed quality checks on incoming datasets by identifying missing values, duplications, and formatting errors prior to model ingestion. Standardized unstructured data records using Excel and basic Python scripts and documented error patterns to update internal SOPs.