Aliyu S.

Key Skills

Software

AWS SageMaker

Data Annotation Tech

CrowdSource

Mercor

Top Subject Matter

LLM training (RLHF)

ASR model fine-tuning

preference labeling

Top Data Types

Text

Audio

Top Task Types

RLHF

Data Collection

Freelancer Overview

Freelance AI Trainer & Dataset Engineer (Remote) — RLHF training datasets, preference labeling, and LLM evaluation. Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal, Proprietary Tooling, and N. Education includes Bachelor of Engineering, Kurdistan University (2026) and Diploma, International Academy for Training and Studies (2021). AI-training focus includes data types such as Text and labeling workflows including RLHF and Data Collection.

Labeling Experience

Freelance AI Trainer & Dataset Engineer (Remote) — RLHF training datasets, preference labeling, and LLM evaluation

TextRLHF

Led RLHF-oriented dataset engineering by preparing preference-labeled and QA-reviewed samples for supervised fine-tuning and LLM evaluation. Built a repeatable annotation pipeline spanning transcription-derived text, preference labeling, and quality assurance to minimize pipeline errors. Evaluated LLM outputs for factual accuracy, response quality, and safety using preference ranking criteria aligned with RLHF methodologies. • Prepared and annotated 1,000+ labeled text samples for supervised fine-tuning. • Applied multi-pass QA reviews to deliver zero-defect datasets. • Implemented preference labeling workflows and quality checks across model cycles. • Performed LLM output evaluation using RLHF-aligned ranking criteria.

2023 - Present

Volunteer Data Collector & QA Analyst — field data validation and error correction

TextData Collection

Conducted field data collection and QA to produce submission-ready datasets for training and review processes. Performed structured collection across multiple sites with cross-field validation and error correction to maintain record consistency. Established annotation discipline through rigorous quality checks that eliminated rejected submissions and dataset inconsistencies. • Collected and validated field data across 5+ sites. • Achieved 100% clean submissions with zero rejected records at official review. • Reduced dataset inconsistencies to zero using systematic QA. • Performed error correction and cross-field validation for annotation readiness.

2022 - 2022

Education

Work History