Marta D.

Key Skills

Software

Scale AI

Remotasks

Top Subject Matter

LLM Evaluation & AI Training Data (RLHF)

Multimodal AI & Computer Vision Annotation

Multilingual NLP & Linguistic Data Quality

Top Data Types

Image

Text

Document

Top Task Types

Segmentation

Bounding Box

Classification

Evaluation/Rating

Prompt + Response Writing (SFT)

Data Collection

Entity (NER) Classification

Object Detection

Transcription

RLHF

Question Answering

Text Generation

Fine-tuning

Freelancer Overview

MultiMango – Aether Project & Outlier | AI Data Annotation and LLM Evaluation Specialist. Brings 2+ years of experience in AI-focused data annotation, multilingual content evaluation, and quality assurance workflows for LLM training. Core strengths include multimodal annotation, reasoning evaluation, factuality review, and linguistic quality assessment across Ukrainian, Polish, Russian, and English. Education includes a Master’s Degree in Data Science from the University of Warsaw. AI training experience includes Image, Text, and Multilingual Data with workflows such as Rating, Validation, Text-to-Image Evaluation, and Multimodal Analysis. Experienced in assessing AI-generated outputs for accuracy, clarity, tone, cultural relevance, and guideline compliance in fast-paced AI training environments.

Labeling Experience

Outlier | LLM Evaluation Specialist & AI Data Annotation

Text

Worked as an LLM Evaluation Specialist performing human evaluation of AI-generated responses. Assessed response accuracy, reasoning quality, clarity, tone, completeness, and factual correctness against conversational and system guidelines. Produced annotated evaluation data highlighting strengths, weaknesses, and factual issues, including for multilingual outputs. • Rated LLM answers to user queries for correctness, reasoning quality, and guideline adherence. • Annotated strengths, weaknesses, and factual inaccuracies to generate training/evaluation data. • Evaluated multilingual responses (Ukrainian, Polish, Russian) for linguistic and cultural appropriateness. • Reviewed and processed multilingual sensitive or policy-restricted content for safety compliance.

2025 - Present

MultiMango – Aether Project | Data Annotation Specialist

ImageSegmentation

Served as a Data Annotation Specialist supporting multimodal AI/ML training across vision and text domains. Produced detailed visual annotations including object-level and fine-grained labeling for small elements and spatial relationships. Verified annotation completeness and correctness against project guidelines and evaluation standards. • Created object-level and fine-grained image labels focusing on subtle visual elements. • Labeled products using structured descriptions and reference images from multiple angles. • Performed vector graphics labeling and documented detection failures by the ML model. • Reviewed face swap and text-to-image annotation outputs for correctness and guideline compliance.

2025 - Present

Education

Work History