Felton

Key Skills

Software

Other

Top Subject Matter

LLM output evaluation

instruction-data quality control

and training sample screening

Top Data Types

Text

Document

Top Task Types

Prompt + Response Writing (SFT)

Data Collection

Freelancer Overview

LLM Training Text Annotator / AI Sample Auditor — grading, correction, and evaluation of LLM outputs. Core strengths include Other. AI-training focus includes data types such as Text and Document and labeling workflows including Evaluation, Rating, and Prompt + Response Writing (SFT).

Labeling Experience

Training Data Cleaning & Archiving — filtering, organizing, and packaging text datasets

OtherDocumentData Collection

They perform cleaning and archiving of text training data by sorting batch dialogue materials and filtering for high-value examples. They remove low-quality or invalid content and package datasets in upload-ready formats. • Filter and remove invalid or low-quality training text • Organize and archive training files (TXT/DOCX) • Prepare lightweight datasets that meet platform format/size limits • Deliver standardized training data for upload and screening

2025 - Present

Standardized Multi-round Review / QA for AI Training Samples

OtherText

They implement a multi-round review mechanism that combines generation, testing, and review using unified evaluation criteria. They conduct cross-review and self-inspection for bulk training materials to sustain labeling accuracy over iterative recheck workflows. • Define unified criteria for consistent evaluation • Use generation-test-review workflow for multi-round checks • Perform cross-review/self-inspection on bulk datasets • Maintain accuracy for mass labeling and recheck cycles

2025 - Present

LLM SFT Fine-tuning Instruction Sample Creator — prompt and sample construction for training data annotation

OtherTextPrompt + Response Writing (SFT)

They construct training instructions and example dialogues for Q&A, code development, and complex reasoning scenarios. They write and refine instruction samples aligned with LLM SFT fine-tuning text specifications to support consistent labeling and training. • Design standardized layered prompts for varied scenarios • Build dialogue samples for Q&A, coding, and reasoning • Write and polish instruction samples for SFT labeling • Ensure training-spec compliance with platform upload requirements

2025 - Present

LLM Training Text Annotator / AI Sample Auditor — grading, correction, and evaluation of LLM outputs

OtherText

They audit and grade LLM-generated outputs for completeness, logical coherence, and factual authenticity. They identify defective samples such as hallucinations, code bugs, and logical gaps, then rewrite or optimize incorrect content for better training quality. • Score AI outputs using unified evaluation criteria • Mark defects and support rewriting of incorrect responses • Locate hallucinations, logical gaps, and coding errors • Improve prompts to iterate higher-quality samples

2025 - Present

Education

Work History