For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
Diego Villarroel

Diego Villarroel

Expert on data labeling in spanish, transcriptions, AI training,

Bolivia flagLa Paz, Bolivia
$6.00/hrEntry LevelCrowdsourceOtherInternal Proprietary Tooling

Key Skills

Software

CrowdSourceCrowdSource
Other
Internal/Proprietary Tooling

Top Subject Matter

Promt plus Reponse Writting
Labeling, training AI with subtitles in spanish and english
Languages and travel planning

Top Data Types

DocumentDocument
TextText

Top Task Types

Prompt Response Writing SFT
Text Summarization

Freelancer Overview

Magister in Business Administration with a C1 Leavel of English I am a bilingual AI Trainer and Data Labeling Specialist with extensive experience in annotation, evaluation, and quality assurance for large language models (LLMs). I have worked on projects involving prompt evaluation, factuality assessment, creative generation, and audio-text alignment across platforms like Outlier, Populii, and Thoth AI. My background includes both hands-on annotation and senior QA review roles, where I’ve ensured that outputs meet accuracy, clarity, and intent-matching standards. I specialize in multilingual tasks (English-Spanish), with proven skills in prompt refinement, conversational analysis, and model alignment. I’ve also contributed to voice/audio evaluation projects, transcribed AI-generated captions, and helped train assistant models in customer support and educational use cases. My ability to adapt to diverse guidelines and deliver high-quality feedback makes me a strong asset to any AI training initiative.

Entry LevelEnglishSpanish

Labeling Experience

Ai Training for prompt and response

Internal Proprietary ToolingTextPrompt Response Writing SFT
On this project I worked in two similar ones, Flamingo and Bulba, in which I am in charge of reviewing the responses of two models, comparing what they answer, witch answer was the best and if the prompt was well written. This project took around 15 cases per day.

On this project I worked in two similar ones, Flamingo and Bulba, in which I am in charge of reviewing the responses of two models, comparing what they answer, witch answer was the best and if the prompt was well written. This project took around 15 cases per day.

2024 - 2024

Spanish ASR Project

Internal Proprietary ToolingTextText SummarizationEvaluation Rating
In this Spanish ASR project, I was in charge of transcribing videos or audios of what was heard, following rules for transcription which went to QA to be graded. This was done on a daily basis depending on the duration of the audios or videos transcribed between 300 to 400 cases.

In this Spanish ASR project, I was in charge of transcribing videos or audios of what was heard, following rules for transcription which went to QA to be graded. This was done on a daily basis depending on the duration of the audios or videos transcribed between 300 to 400 cases.

2023 - 2023

Untitled labeling experience

No description provided.

No description provided.

Not specified

Education

S

Salamanca Univeristy

Mastery in Business Administration, Business Administration

Mastery in Business Administration
2021 - 2022
U

Universidad Privada Boliviana

Bachelor's Degree in Business Administration, Business Administration

Bachelor's Degree in Business Administration
2013 - 2018

Work History

O

Outlier

AI Trainer & QA Reviewer

La Paz
2024 - 2025
T

Thoth AI

Labeler

Singapore
2023 - 2024