For employers

Hire this AI Trainer

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

LoginCreate account
Invite to Job
M
Marcela Thesz

Marcela Thesz

AI Evaluator / Data Annotatot

Argentina flagN/A, Argentina
$15.00/hrEntry LevelLabelboxSuperannotateToloka

Key Skills

Software

LabelboxLabelbox
SuperAnnotateSuperAnnotate
TolokaToloka

Top Subject Matter

Conversational AI
Rioplatense Spanish
Emotional Intelligence

Top Data Types

TextText
DocumentDocument
ImageImage

Top Task Types

Text GenerationText Generation
RLHFRLHF
Red TeamingRed Teaming

Freelancer Overview

I am a freelance AI Trainer and bilingual SME (EN/ES) with over 20 years of experience in human communication, learning design, and editorial production. I currently work in RLHF and Safety & Trust evaluation streams, designing per-turn rubrics, ranking LLM responses, and assessing model outputs for harmful content, bias, and policy compliance. I also specialize in Emotional Intelligence evaluation in Rioplatense Spanish, bringing a culturally grounded and linguistically precise perspective to AI output assessment.

Entry LevelEnglishSpanish

Labeling Experience

AI Trainer (Freelance) | SME Careers

TextRLHF
I evaluated and ranked LLM responses across diverse domains to ensure high-quality model alignment. Safety and Trust evaluations were conducted to audit outputs for harmful content and biases. RLHF quality rubrics were applied to improve logic and helpfulness of language models. • Ranked model responses using specialized rating systems. • Conducted safety audits for bias and harmful output detection. • Aligned AI outputs with human values and strict policy standards. • Used bilingual expertise for EN/ES content evaluation.

I evaluated and ranked LLM responses across diverse domains to ensure high-quality model alignment. Safety and Trust evaluations were conducted to audit outputs for harmful content and biases. RLHF quality rubrics were applied to improve logic and helpfulness of language models. • Ranked model responses using specialized rating systems. • Conducted safety audits for bias and harmful output detection. • Aligned AI outputs with human values and strict policy standards. • Used bilingual expertise for EN/ES content evaluation.

2026 - Present

AI Evaluator (Freelance) | Alignerr

TextRLHF
I designed evaluation rubrics and rated LLM model responses for emotionally rich conversation prompts in Rioplatense Spanish. This work required assessing emotional intelligence, tone, and cultural appropriateness across multi-turn interactions. The role included side-by-side (SxS) ranking of model outputs and identifying best fits for human-AI alignment. • Developed per-turn evaluation rubrics for dialogue annotation. • Evaluated LLM responses for emotional content and linguistic nuance. • Conducted quality control to ensure culturally appropriate outputs. • Used structured feedback to improve response selection.

I designed evaluation rubrics and rated LLM model responses for emotionally rich conversation prompts in Rioplatense Spanish. This work required assessing emotional intelligence, tone, and cultural appropriateness across multi-turn interactions. The role included side-by-side (SxS) ranking of model outputs and identifying best fits for human-AI alignment. • Developed per-turn evaluation rubrics for dialogue annotation. • Evaluated LLM responses for emotional content and linguistic nuance. • Conducted quality control to ensure culturally appropriate outputs. • Used structured feedback to improve response selection.

2026 - Present

Education

N

N/A

Specialization, Mindfulness and Secular Ethics

Specialization
Not specified
U

Universidad Nacional de Córdoba

Bachelor of Arts, Communications and Media

Bachelor of Arts
Not specified

Work History

M

marcelathesz.com

Lead Content Producer and Author

N/A
2015 - Present
I

Independent Projects

Digital Production Manager

N/A
2020 - 2025