For employers

Hire this AI Trainer

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

Invite to Job
M

Marin Beštek

AI Data Specialist – Content Evaluation, Media Annotation & Sports Expert

CROATIA flag
Varaždin, Croatia
$25.00/hrIntermediateOtherAxiom AIAppen

Key Skills

Software

Other
Axiom AI
AppenAppen
Data Annotation TechData Annotation Tech
OneFormaOneForma
MindriftMindrift
Mighty AIMighty AI
RemotasksRemotasks

Top Subject Matter

Generative AI
Sports & Fitness
Digital Media & Video Production

Top Data Types

ImageImage
TextText
DocumentDocument

Top Task Types

Classification

Freelancer Overview

Experienced AI Data Specialist with a strong background in content evaluation, prompt engineering, and multi-modal data labeling. Over the past 2+ years, I have specialized in fine-tuning LLMs and image generation models for digital media projects, focusing on classification, sentiment analysis, and fact-checking to ensure model accuracy and safety. My professional foundation includes a High School Diploma from the Varaždin School of Economics (2020), which equipped me with strong administrative and research skills. Beyond technology, I bring deep domain expertise in sports and fitness as an active practitioner, providing a unique perspective for specialized recreation and athletic data projects. I am a native Croatian speaker with C1 English proficiency, highly disciplined and accustomed to high-precision remote workflows.

IntermediateCroatianEnglish

Labeling Experience

AI Content Synthesis & Media Evaluation

TextClassification
Managed a large-scale data annotation and content synthesis project for digital media channels. The scope involved evaluating AI-generated scripts for logical consistency, factual accuracy, and linguistic flow (RLHF). Specific tasks included text categorization, sentiment analysis, and multi-modal data labeling (pairing AI-generated voiceovers with visual assets). I adhered to strict quality control measures, identifying model 'hallucinations' and refining prompts to achieve a 95%+ accuracy rate in final outputs. The project required handling thousands of data points monthly, ensuring high-quality training data for automated video production workflows.

Managed a large-scale data annotation and content synthesis project for digital media channels. The scope involved evaluating AI-generated scripts for logical consistency, factual accuracy, and linguistic flow (RLHF). Specific tasks included text categorization, sentiment analysis, and multi-modal data labeling (pairing AI-generated voiceovers with visual assets). I adhered to strict quality control measures, identifying model 'hallucinations' and refining prompts to achieve a 95%+ accuracy rate in final outputs. The project required handling thousands of data points monthly, ensuring high-quality training data for automated video production workflows.

2023 - Present

Data Annotation – labeling, tagging, and content evaluation

OtherImageClassification
I have practical experience in data annotation, labeling, tagging, and content evaluation for generative AI image and content creation tools. This involved classifying images and evaluating outputs for accuracy and quality. I utilized a high level of precision and attention to detail in every annotation task performed. • Labeled and annotated images for training data used in generative AI models. • Tagged content according to established guidelines to improve model accuracy and relevance. • Evaluated and provided feedback on generated content, identifying quality issues and inconsistencies. • Adapted to new digital platforms and software tools quickly and efficiently.

I have practical experience in data annotation, labeling, tagging, and content evaluation for generative AI image and content creation tools. This involved classifying images and evaluating outputs for accuracy and quality. I utilized a high level of precision and attention to detail in every annotation task performed. • Labeled and annotated images for training data used in generative AI models. • Tagged content according to established guidelines to improve model accuracy and relevance. • Evaluated and provided feedback on generated content, identifying quality issues and inconsistencies. • Adapted to new digital platforms and software tools quickly and efficiently.

Not specified

Education

V

Varaždin School of Economics

High School Diploma, Economics

High School Diploma
2016 - 2020
N

Novi Marof Elementary School

Elementary School Diploma, General Education

Elementary School Diploma
2008 - 2016

Work History

D

Dynamic Trade

Assistant Service Advisor

Varaždin
2025 - Present
I

Inter Cars

Warehouse & Logistics Associate

Varaždin
2025 - 2025