Isaiah A.

Key Skills

Software

Appen

Toloka

Clickworker

OneForma

Top Subject Matter

LLM response evaluation

RLHF preference data

content safety and quality moderation

Top Data Types

Text

Image

Document

Top Task Types

Data Collection

Freelancer Overview

I possess hands on experience in AI training data annotation, text evaluation, and content moderation across leading platforms such as Appen, Toloka AI, Clickworker, and OneForma. My work has involved evaluating AI generated responses for accuracy, coherence, relevance, tone, and safety using structured annotation guidelines and multi dimensional scoring systems. I have practical experience with RLHF preference ranking, dataset validation, prompt analysis, and content review workflows designed to improve large language model performance. Across these projects, I consistently maintained high quality standards while working independently in remote environments with strict deadlines and calibration requirements. What distinguishes me is my strong analytical background from both Law and English Language studies, which has sharpened my critical reasoning, attention to detail, and written communication abilities. I am skilled at identifying hallucinations, bias, harmful content, formatting inconsistencies, and low quality data entries while providing clear written justifications for annotation decisions. In addition to AI text evaluation, I have handled search quality assessment, transcription validation, multilingual data contribution, web research, and content categorisation tasks. My ability to combine precision, linguistic understanding, and structured evaluation enables me to contribute effectively to high accuracy AI training and data labeling projects.

Labeling Experience

AI Text Evaluator & Data Annotator (Appen)

AppenText

Provided multi-dimensional scoring of AI-generated text for accuracy, relevance, coherence, tone, and safety. Conducted RLHF pairwise preference comparisons by ranking two AI outputs side by side to generate human feedback training signals. Identified hallucinations, factual errors, bias, harmful content, and formatting inconsistencies in annotation queues. Included written justifications for every rating decision according to platform rubrics. • Rated AI responses using structured scoring rubrics • Performed RLHF preference ranking (pairwise comparisons) • Flagged harmful or incorrect content and formatting issues • Wrote justifications to support model improvement

2024 - 2025

AI Content Reviewer (Toloka AI)

TolokaText

Performed NLP dataset annotation, categorization, and moderation-focused evaluation of AI outputs. Identified sensitive, misleading, and policy-violating material as part of content moderation workflows. Validated and cleaned datasets by finding duplicates, corrupted entries, and low-quality labels. • Annotated and categorized diverse text datasets for NLP • Evaluated outputs for policy-violating and sensitive content • Validated and cleaned datasets (duplicates/corruption/low-quality) • Produced consistent high-accuracy results remotely

2023 - 2025

Freelance Data Annotator (Clickworker)

ClickworkerText

Performed freelance text creation, web research, image tagging, and data categorization tasks across multiple microtask projects. Completed UHRS tasks including search quality evaluation and ad relevance rating. Applied consistent labeling standards to maintain accuracy scores above platform benchmarks. • Completed search quality evaluation and ad relevance rating (UHRS) • Performed text creation and data categorization • Applied consistent labeling standards in high-volume queues • Managed independent workload and met deadlines/quality thresholds

2022 - 2024

AI Data Contributor (OneForma/Centific)

OneFormaTextData Collection

Contributed to multilingual data collection and text annotation projects supporting AI and NLP model development. Conducted transcription and translation validation plus linguistic data quality checks for English datasets. Evaluated search engine results and web content for relevance, quality, and user intent alignment while adhering to strict project guidelines. • Performed transcription and translation validation • Conducted linguistic data quality checks • Evaluated search results/web content for relevance and intent • Followed confidentiality and project-specific guidelines

2022 - 2023

Education

Work History