Biruk G.

Key Skills

Software

Appen

OneForma

Top Subject Matter

AI Training & Model Evaluation (RLHF)

Civil Engineering & Structural Design

Technical Fact-Checking & Constraint Verification

Top Data Types

Text

Document

Top Task Types

Segmentation

Object Detection

Text Generation

Fine-tuning

Freelancer Overview

Real Estate Project Manager and Co-Owner in Contract Review, Compliance, and Legal Research. Brings 19+ years of professional experience across complex professional workflows, research, and quality-focused execution. Experienced AI Data Trainer and Model Evaluator specializing in technical domain auditing. Expert in RLHF workflows, structural engineering fact-checking, and checking AI outputs against rigorous logic constraints. Skilled in identifying subtle model failures, verifying structural metrics, and writing objective final evaluation verdicts to improve LLM accuracy. Education includes Undergraduate Studies, Self-Funded Higher Education Program (2020) and Bachelor of Science, Hawassa University (2008).

Labeling Experience

Multimodal Language Model Evaluator & Quality Assurance Auditor

ImageSegmentation

Evaluated and audited multimodal large language model (LLM) outputs for high-precision Handwriting Synthesis and Refinement datasets. Reviewed complex visual data across diverse language scripts (including localized character sets and ligatures) to identify micro-glitches, alignment shifts, and stroke anatomy defects. Validated model outputs against strict quality matrices, managing edge cases in image clarity, visual legibility, and aesthetic consistency to refine computer vision training benchmarks.

2026 - Present

Multilingual Cultural Content & Image-Text Validator

ImageText Generation

Specialized in validating, rewriting, and auditing complex multimodal image-text paired datasets for localized cultural accuracy. Experienced in evaluating high-context visual elements (including traditional attire, architecture, regional cuisine, and festivals) and aligning them with precise native-language captions and English descriptions. Focused on tracking mislabeled media, resolving translation ambiguities, and adding critical regional context to train culturally aware AI models.

2026 - Present

AI Data Annotator & LLM Evaluation Specialist

TextEvaluation/Rating

Conducted rigorous Side-by-Side (SBS) comparative analysis of LLM response profiles to optimize conversational and factual accuracy. Evaluated model adherence to strict persona guidelines, reasoning structures, and technical parameters. Successfully cleared advanced platform certifications with an 87.5% performance rating, demonstrating deep familiarity with auditing model hallucinations, factual validation, and safety guardrails.

2026 - Present

Education

Work History