Gabriel Lemos

Key Skills

Software

Mindrift

Other

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Text

Top Label Types

Classification

Computer Programming Coding

Evaluation Rating

Function Calling

Prompt Response Writing SFT

Question Answering

RLHF

Text Generation

Text Summarization

Freelancer Overview

I am a Software Engineer and Computer Science student with specialized experience in enhancing the reasoning and performance of state-of-the-art AI models. Currently, I serve as an AI Agent Evaluation Analyst and AI Trainer, where I leverage RLHF (Reinforcement Learning from Human Feedback) and RLVR (Reinforcement Learning from Verifiable Reward) protocols to train models in complex algorithmic generation and verifiable reasoning. My work involves defining technical scoring criteria, performing corrective adjustments to eliminate errors, and auditing AI-generated code to ensure it meets production-grade standards. Beyond data labeling and model alignment, I bring a strong technical background in full-stack development, including proficiency in React, Kotlin, and Python. I have a proven track record of implementing the Model Context Protocol (MCP) and refining tool-calling sequences to bridge the gap between autonomous agents and external API ecosystems. This combination of engineering logic and experience in advanced AI training methodologies allows me to optimize agent reliability and performance across diverse task execution environments.

Labeling Experience

AI Trainer / Code Reviewer

OtherTextText GenerationRLHF

As an AI Trainer and Code Reviewer at Outlier, I specialize in high-level data labeling through the auditing and optimization of AI-generated code from various SOTA models. My work focuses on advanced model alignment using RLHF and RLVR protocols to enhance complex algorithmic generation and verifiable reasoning. To ensure production-grade quality, I implement the Model Context Protocol (MCP) and tool-calling to bridge the gap between AI models and external API ecosystems.

As an AI Trainer and Code Reviewer at Outlier, I specialize in high-level data labeling through the auditing and optimization of AI-generated code from various SOTA models. My work focuses on advanced model alignment using RLHF and RLVR protocols to enhance complex algorithmic generation and verifiable reasoning. To ensure production-grade quality, I implement the Model Context Protocol (MCP) and tool-calling to bridge the gap between AI models and external API ecosystems.

2025

AI Agent Evaluation Analyst

MindriftTextClassificationRLHF

As a remote AI Agent Evaluation Analyst at Mindrift , I optimize autonomous AI agents by evaluating their performance across behavior modeling and MCP task execution environments. My work involves specialized data labeling tasks, including defining technical scoring criteria and performing corrective RLHF-based adjustments to eliminate errors and enhance reliability. I leverage prompt engineering to refine agent reasoning, specifically ensuring successful tool-calling sequences and persistent database state management to maintain high-quality, production-grade agent performance.

As a remote AI Agent Evaluation Analyst at Mindrift , I optimize autonomous AI agents by evaluating their performance across behavior modeling and MCP task execution environments. My work involves specialized data labeling tasks, including defining technical scoring criteria and performing corrective RLHF-based adjustments to eliminate errors and enhance reliability. I leverage prompt engineering to refine agent reasoning, specifically ensuring successful tool-calling sequences and persistent database state management to maintain high-quality, production-grade agent performance.

2025

Education

Work History