Lucia Mendoza

Key Skills

Software

Google Cloud Vertex AI

Labelbox

Scale AI

Top Subject Matter

LLM evaluation in Spanish

Sentiment Analysis

Object Detection and Classification

Top Data Types

Computer Code Programming

Image

Text

Top Task Types

Computer Programming Coding

Emotion Recognition

Prompt Response Writing SFT

Freelancer Overview

I am an experienced AI Data Labeling Specialist with 1 to 3 years of hands-on experience in data annotation and AI training. My expertise spans across multiple domains, including text, image, and video data labeling, with a strong focus on precision and quality. I have worked extensively with industry-leading tools such as Labelbox and Scale AI, applying advanced labeling techniques like bounding box annotation, entity classification (NER), and text summarization. Additionally, I have experience verifying and refining AI-generated responses related to programming topics, ensuring that the outputs are accurate, relevant, and aligned with best practices. My ability to meticulously label and annotate data, combined with my experience in AI prompt verification, ensures that AI models are trained with high-quality datasets, leading to more accurate and reliable outcomes. I am passionate about contributing to innovative AI projects and have a keen eye for detail, which sets me apart in ensuring data accuracy and consistency.

Labeling Experience

Code Turns Training

Scale AIComputer Code ProgrammingEvaluation RatingComputer Programming Coding

The Flamingo Code Turns project involved generating and evaluating AI-driven programming prompts designed to challenge AI models across different levels of expertise, from Junior to Senior developers. The project aimed to refine the model's ability to generate and assess code quality, as well as to ensure that it could produce solutions that meet industry standards. The process involved creating diverse programming prompts covering a range of topics, such as algorithms, data structures, and software engineering principles. Each prompt was designed to test the AI model’s coding abilities and decision-making process. For every prompt, the AI model generated two distinct responses. My role was to evaluate these responses by: Testing the Code: I executed the provided code to ensure functionality, correctness, and adherence to best practices. Choosing the Better Response: After thorough testing, I selected the superior response based on factors such as code efficiency, readability, etc.

The Flamingo Code Turns project involved generating and evaluating AI-driven programming prompts designed to challenge AI models across different levels of expertise, from Junior to Senior developers. The project aimed to refine the model's ability to generate and assess code quality, as well as to ensure that it could produce solutions that meet industry standards. The process involved creating diverse programming prompts covering a range of topics, such as algorithms, data structures, and software engineering principles. Each prompt was designed to test the AI model’s coding abilities and decision-making process. For every prompt, the AI model generated two distinct responses. My role was to evaluate these responses by: Testing the Code: I executed the provided code to ensure functionality, correctness, and adherence to best practices. Choosing the Better Response: After thorough testing, I selected the superior response based on factors such as code efficiency, readability, etc.

2024 - 2024

Spanish Calibration

LabelboxTextText GenerationEmotion Recognition

I also engaged in Prompt + Response Writing to generate and verify AI-generated responses based on customer inquiries, ensuring relevance and accuracy. My role included the Evaluation of AI outputs to assess their alignment with sentiment classification goals, followed by detailed feedback to refine the models.

I also engaged in Prompt + Response Writing to generate and verify AI-generated responses based on customer inquiries, ensuring relevance and accuracy. My role included the Evaluation of AI outputs to assess their alignment with sentiment classification goals, followed by detailed feedback to refine the models.

2024 - 2024

Education

Work History