Enrique Perez

Key Skills

Software

Data Annotation Tech

Labelbox

Top Subject Matter

Mathematical and logical resolution

LLM EVALUATION AND TRAINING IN SPANISH

Renewable Energy and Sustainability

Reinforcement Learning with Human Feedback

AI Training for Mathematics, Logical Reasoning, and Engineering Applications

Top Data Types

3D Sensor

Computer Code Programming

Document

Top Task Types

Classification

Computer Programming Coding

Prompt Response Writing SFT

RLHF

Text Summarization

Freelancer Overview

I have been working for the past three months, and I continue to actively contribute, on data annotation platforms such as Data Annotation Tech and Outlier. At Outlier, I am involved in impactful projects like Argon, Merlin, and Kepler, where I evaluate and refine AI systems across various domains. My expertise lies in tasks that require advanced mathematical reasoning, problem-solving, and logical analysis, particularly in practical fields such as electricity, energy systems, and physics. I am committed to delivering precise and high-quality annotations that significantly enhance AI performance in these areas.

Labeling Experience

Merlin

Data Annotation TechTextRLHFEvaluation Rating

The project involves evaluating and refining the performance of AI models by posing problems and questions that challenge their logical, mathematical, and practical reasoning capabilities. For each task, responses are generated by 2 different AI models, which I meticulously review, compare, and modify as needed to ensure accuracy, clarity, and alignment with the task's requirements. The scope includes fine-tuning responses to achieve optimal quality, contributing to the development of highly effective AI systems. Strict quality measures are adhered to, focusing on precision, relevance, and usability, ensuring that the final output meets the highest standards. in this project it is mandatory to make at least 3 turns on the same subject.

The project involves evaluating and refining the performance of AI models by posing problems and questions that challenge their logical, mathematical, and practical reasoning capabilities. For each task, responses are generated by 2 different AI models, which I meticulously review, compare, and modify as needed to ensure accuracy, clarity, and alignment with the task's requirements. The scope includes fine-tuning responses to achieve optimal quality, contributing to the development of highly effective AI systems. Strict quality measures are adhered to, focusing on precision, relevance, and usability, ensuring that the final output meets the highest standards. in this project it is mandatory to make at least 3 turns on the same subject.

2024 - 2024

Outlier Kepler RLHF

Don T DiscloseTextRLHF

In the Kepler project at Outlier, I focus exclusively on advanced mathematical tasks spanning various types and levels of complexity. My primary objective is to challenge the AI models by designing and presenting problems that expose potential weaknesses in their reasoning or computation. This process often requires multiple iterations, where I analyze the model's responses to identify points of failure. Once a failure is detected, I meticulously correct the error and refine the response to achieve an optimal, flawless solution. This iterative approach ensures that the model not only improves its mathematical reasoning but also aligns with high-quality standards, delivering precise and accurate outputs.

In the Kepler project at Outlier, I focus exclusively on advanced mathematical tasks spanning various types and levels of complexity. My primary objective is to challenge the AI models by designing and presenting problems that expose potential weaknesses in their reasoning or computation. This process often requires multiple iterations, where I analyze the model's responses to identify points of failure. Once a failure is detected, I meticulously correct the error and refine the response to achieve an optimal, flawless solution. This iterative approach ensures that the model not only improves its mathematical reasoning but also aligns with high-quality standards, delivering precise and accurate outputs.

2024

Argon

Data Annotation TechTextSegmentationRLHF

The project involves evaluating and refining the performance of AI models by posing problems and questions that challenge their logical, mathematical, and practical reasoning capabilities. For each task, responses are generated by 2-3 different AI models, which I meticulously review, compare, and modify as needed to ensure accuracy, clarity, and alignment with the task's requirements. The scope includes fine-tuning responses to achieve optimal quality, contributing to the development of highly effective AI systems. Strict quality measures are adhered to, focusing on precision, relevance, and usability, ensuring that the final output meets the highest standards.

The project involves evaluating and refining the performance of AI models by posing problems and questions that challenge their logical, mathematical, and practical reasoning capabilities. For each task, responses are generated by 2-3 different AI models, which I meticulously review, compare, and modify as needed to ensure accuracy, clarity, and alignment with the task's requirements. The scope includes fine-tuning responses to achieve optimal quality, contributing to the development of highly effective AI systems. Strict quality measures are adhered to, focusing on precision, relevance, and usability, ensuring that the final output meets the highest standards.

2024

Education

Work History