Aminu K.

Key Skills

Software

Mercor

Other

Top Subject Matter

LLM response evaluation within RLHF pipelines for instruction-following and safety alignment

Text and dialogue annotation for AI/ML training and fine-tuning

NLP labeling (NER, sentiment, intent) with QA and ethical guidelines

Top Data Types

Text

Image

Top Task Types

Fine-tuning

Entity (NER) Classification

Freelancer Overview

AI Response Evaluator, Project Meereen (via Mercor). Brings 8+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Mercor and Atlas. Education includes Bachelor of Technology, Abubakar Tafawa Balewa University (ATBU) (2023). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Fine-tuning.

Labeling Experience

Data Labeler, Atlas

TextEntity (NER) Classification

Labeled structured and unstructured datasets for NLP and computer vision tasks, including named entity recognition (NER), sentiment analysis, and intent classification. Followed comprehensive labeling schemas and resolved inter-annotator disagreements using escalation protocols and documentation. Participated in quality assurance reviews by flagging inconsistencies and providing feedback to improve labeling pipeline efficiency. • Applied NER, sentiment, and intent classification labeling schemas. • Used escalation and documentation to resolve labeling disagreements. • Flagged inconsistencies during QA and improved pipeline efficiency through feedback. • Handled sensitive and complex annotation scenarios with ethical AI principles.

2026 - Present

AI Annotation Specialist, Mercor

MercorTextFine-tuning

Performed multi-label and binary classification annotation tasks to support training and fine-tuning of AI/ML models. Annotated text, dialogue, and instruction-response pairs according to detailed annotation guidelines and quality benchmarks. Collaborated with remote annotation teams and quality reviewers to resolve ambiguous cases and maintain labeling standards. • Conducted binary and multi-label classification for model training datasets. • Labeled instruction-response and dialogue data with guideline adherence. • Resolved ambiguous cases via collaboration with reviewers and escalation. • Met throughput targets while sustaining high accuracy to improve dataset quality.

2026

AI Response Evaluator, Project Meereen (via Mercor)

MercorText

Evaluated AI-generated responses as part of an RLHF (Reinforcement Learning from Human Feedback) pipeline, grading outputs for accuracy, helpfulness, harmlessness, and alignment with human preferences. Applied detailed rubrics to assess multiple dimensions including factual correctness, tone, coherence, and instruction-following. Provided structured written justifications identifying edge cases, hallucinations, and policy violations to guide model improvement. • Used rubric-based judgment to rate model outputs consistently. • Performed edge case identification and hallucination/policy violation detection. • Maintained annotation quality and inter-rater reliability in large-scale evaluation tasks. • Delivered structured feedback for downstream model improvement.

2026

Education

Work History