Cassondra Phillips

Key Skills

Software

Appen

Clickworker

Lionbridge

OneForma

Remotasks

Telus

Scale AI

Other

Top Subject Matter

Large Language Models (LLMs)

AI-generated content

Social Media Content Moderation

Top Data Types

Computer Code Programming

Text

Video

Audio

Document

Top Task Types

Audio Recording

Data Collection

Prompt Response Writing SFT

Question Answering

Text Generation

RLHF

Classification

Freelancer Overview

LLM Annotator. Brings 11+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other and Telus. Education includes Master of Science, Southern New Hampshire University (2024) and Associate of Science, Hazard Community and Technical College (2015). AI-training focus includes data types such as Text and labeling workflows including RLHF, Classification, and Evaluation.

Labeling Experience

LLM Annotator

OtherTextRLHF

As an LLM Annotator, I labeled diverse datasets for large language model training and evaluation. I assessed AI-generated content for coherence, accuracy, safety, and ethics in alignment with strict project guidelines. I also provided structured feedback to support Reinforcement Learning from Human Feedback (RLHF) and maintained rigorous annotation standards. • Evaluated outputs for summarization, question answering, and code generation. • Collaborated to clarify ambiguous cases and ensured guideline updates. • Documented annotation procedures to support inter-annotator agreement. • Upheld quality benchmarks and project service level agreements.

As an LLM Annotator, I labeled diverse datasets for large language model training and evaluation. I assessed AI-generated content for coherence, accuracy, safety, and ethics in alignment with strict project guidelines. I also provided structured feedback to support Reinforcement Learning from Human Feedback (RLHF) and maintained rigorous annotation standards. • Evaluated outputs for summarization, question answering, and code generation. • Collaborated to clarify ambiguous cases and ensured guideline updates. • Documented annotation procedures to support inter-annotator agreement. • Upheld quality benchmarks and project service level agreements.

2025 - 2026

Project Arrow

AppenTextEvaluation Rating

Involves evaluating responses generated by large language models (LLMs) for accuracy, coherence, and relevance based on user queries.

Involves evaluating responses generated by large language models (LLMs) for accuracy, coherence, and relevance based on user queries.

2022 - 2024

Project Yukon

AppenTextEvaluation Rating

Involves evaluating search engine results to ensure relevance and quality, focusing on the user's intent behind the query.

Involves evaluating search engine results to ensure relevance and quality, focusing on the user's intent behind the query.

2021 - 2024

Project Flamingo

Scale AITextPrompt Response Writing SFT

Evaluated questions and answers provided by a LLM using very detailed instructions.

Evaluated questions and answers provided by a LLM using very detailed instructions.

2023 - 2023

Risk Analyst

OtherTextClassification

As a Risk Analyst, I reviewed and classified social media content to determine safety, appropriateness, and policy compliance. I flagged potential threats such as hate speech and abuse following platform and client protocols. My work contributed to strategic improvements in risk mitigation and content moderation. • Classified content for risk types and policy adherence. • Escalated posts for further action and guideline review. • Identified brand and reputational risk cases. • Supported ongoing improvement in classification accuracy.

As a Risk Analyst, I reviewed and classified social media content to determine safety, appropriateness, and policy compliance. I flagged potential threats such as hate speech and abuse following platform and client protocols. My work contributed to strategic improvements in risk mitigation and content moderation. • Classified content for risk types and policy adherence. • Escalated posts for further action and guideline review. • Identified brand and reputational risk cases. • Supported ongoing improvement in classification accuracy.

2018 - 2020

Education

Work History