For employers

Hire this AI Trainer

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

Invite to Job
Marcia Coelho

Marcia Coelho

AI Data & Language Services Specialist - Media & Communication

NETHERLANDS flag
Amsterdam, Netherlands
$18.00/hrIntermediateAppenClickworkerCrowdsource

Key Skills

Software

AppenAppen
ClickworkerClickworker
CrowdSourceCrowdSource
iMeritiMerit
LabelboxLabelbox
MercorMercor
SuperAnnotateSuperAnnotate
TelusTelus
Other

Top Subject Matter

No subject matter listed

Top Data Types

AudioAudio
ImageImage
TextText
VideoVideo

Top Label Types

Bounding Box
Classification
Emotion Recognition
Evaluation Rating
Object Detection
Prompt Response Writing SFT
Transcription
Translation Localization

Freelancer Overview

I am an experienced media and communication professional with a strong focus on AI data services, specializing in transcription, data annotation, and labeling for large language models and computer vision projects. My hands-on work includes image, video, and 3D annotation to support multimodal AI systems, as well as data collection and search relevance for large-scale datasets in distributed, international teams. I am skilled in ensuring data quality and consistency, following detailed guidelines for machine learning pipelines, and adapting content through translation and transcreation. My background in media production and digital content creation, combined with advanced language skills and a keen eye for detail, enables me to deliver high-quality training data for a variety of AI applications.

IntermediateEnglishSpanishFrenchPortuguese

Labeling Experience

Audio TTS Evaluation

OtherAudioEvaluation Rating
This is a recurring Audio TTS (Text-to-Speech) Evaluation project in which I am responsible for reviewing and assessing batches of synthesized speech audio files. Each batch contains approximately 100 audio samples, and my role is to evaluate them across multiple qualitative and technical dimensions. For each file, I assess key aspects such as: Overall audio quality Naturalness of the voice Intelligibility and listening effort Pronunciation accuracy Prosodic appropriateness (intonation, rhythm, and stress) I assign scores for each criterion, calculate the average score per file, and provide detailed written feedback highlighting strengths, issues, and areas for improvement. This helps ensure the TTS system meets high standards of clarity, realism, and usability for end users.

This is a recurring Audio TTS (Text-to-Speech) Evaluation project in which I am responsible for reviewing and assessing batches of synthesized speech audio files. Each batch contains approximately 100 audio samples, and my role is to evaluate them across multiple qualitative and technical dimensions. For each file, I assess key aspects such as: Overall audio quality Naturalness of the voice Intelligibility and listening effort Pronunciation accuracy Prosodic appropriateness (intonation, rhythm, and stress) I assign scores for each criterion, calculate the average score per file, and provide detailed written feedback highlighting strengths, issues, and areas for improvement. This helps ensure the TTS system meets high standards of clarity, realism, and usability for end users.

2025

Bounding Box Annotation

OtherImageBounding Box
This project focused on UX Interface Bounding Box Annotation across different operating systems. I was responsible for creating realistic usage scenarios and working with multiple desktop environments, including macOS, Windows, and Ubuntu (Linux). My tasks included: Designing scenarios that reflect real user interactions on desktop interfaces Identifying and annotating all UX elements (icons, menus, windows, buttons, taskbars, docks, etc.) Drawing precise bounding boxes around each interface component Assigning each element a clear title and descriptive label explaining its function and context The goal of the project was to generate high-quality labeled data for training and evaluating computer vision and UI understanding models.

This project focused on UX Interface Bounding Box Annotation across different operating systems. I was responsible for creating realistic usage scenarios and working with multiple desktop environments, including macOS, Windows, and Ubuntu (Linux). My tasks included: Designing scenarios that reflect real user interactions on desktop interfaces Identifying and annotating all UX elements (icons, menus, windows, buttons, taskbars, docks, etc.) Drawing precise bounding boxes around each interface component Assigning each element a clear title and descriptive label explaining its function and context The goal of the project was to generate high-quality labeled data for training and evaluating computer vision and UI understanding models.

2025 - 2025

Education

I

International University of Applied Sciences – IU

Bachelor of Science, Cybersecurity

Bachelor of Science
2025 - 2025
A

Alliance Française de Rio de Janeiro

Certificate, French Language and Culture

Certificate
2014 - 2014

Work History

S

Self-Employed

Media and Communication Specialist

Amsterdam
2015 - 2023
S

Self-Employed

Photographer

Amsterdam
2015 - 2023