Charles Castillo

Key Skills

Software

Appen

Clickworker

Data Annotation Tech

Labelbox

OneForma

Other

Top Subject Matter

Safety Evaluation

AI Chatbots

Research Evaluation

Top Data Types

Document

Image

Text

Top Task Types

Classification

Evaluation Rating

Prompt Response Writing SFT

Question Answering

Text Generation

Freelancer Overview

I am a detail oriented individual specialized analysis and evaluation of large language models. I have worked to identify and correct hallucinations presented in AI generated responses, with a primary focus on writing, mathematics, ethics and law through prompt creation, research, fact checking, and direct editing of responses.

Labeling Experience

Image Recognition

Data Annotation TechImageBounding BoxQuestion Answering

This project focused on testing an AI's facial and object recognition capabilities. A model was provided an image and a set of questions with the understanding that image provided context to the conversation. The models were rated for safety, accuracy, relevance, and honesty. Images varied from photographs, to illustrations, and infographics.

This project focused on testing an AI's facial and object recognition capabilities. A model was provided an image and a set of questions with the understanding that image provided context to the conversation. The models were rated for safety, accuracy, relevance, and honesty. Images varied from photographs, to illustrations, and infographics.

2023 - 2024

Chatbot Evaluation

Data Annotation TechTextText GenerationPrompt Response Writing SFT

This project involved conducting various conversations with an AI. These conversations addressed finance, idle chat, legal advice, life advice, writing, and other queries. Throughout the project, hallucinations were identified and corrected, as well as potential safety concerns relating to ethics and regulations. An emphasis was placed on responses which reflected a natural interaction between two people.

This project involved conducting various conversations with an AI. These conversations addressed finance, idle chat, legal advice, life advice, writing, and other queries. Throughout the project, hallucinations were identified and corrected, as well as potential safety concerns relating to ethics and regulations. An emphasis was placed on responses which reflected a natural interaction between two people.

2023 - 2024

Education

Work History