For employers

Hire this AI Trainer

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

LoginCreate account
Invite to Job
Danylo Zhuravlov

Danylo Zhuravlov

Oracle-Tier AI Trainer - AI & Linguistics

POLAND flag
Wroclaw, Poland
$40.00/hrIntermediateInternal Proprietary Tooling

Key Skills

Software

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

AudioAudio
TextText
Computer Code ProgrammingComputer Code Programming

Top Label Types

Emotion Recognition
Audio Recording
Evaluation Rating
Diagnosis
RLHF
Classification
Fine Tuning
Computer Programming Coding
Prompt Response Writing SFT

Freelancer Overview

I am an Oracle-Tier AI Trainer and data annotation specialist with over 7 years of engineering experience and a Master’s degree in Applied Mathematics. My expertise spans high-precision data labeling, RLHF, SFT, and prompt engineering for large language models, with a proven track record of >98% QA accuracy. I have contributed to advanced AI training workflows, including mathematical reasoning, code evaluation (Python, C++, Lua, SQL), and cross-lingual audio transcription and labeling (UA, RU, PL, EN). My work includes generating and evaluating complex datasets for NLP, voice synthesis, and instruction following, as well as building robust automation frameworks and conducting detailed root cause analysis for model improvement. I leverage my technical background and multilingual skills to deliver reliable, high-quality training data for AI and machine learning applications.

IntermediateRussianUkrainianPolishEnglish

Labeling Experience

Oracle-Tier AI Trainer & Multidisciplinary Subject Matter Expert

Internal Proprietary ToolingTextDiagnosisRLHF
CORE DOMAINS & RESPONSIBILITIES: 🚀 1. Advanced RLHF & SFT (Model Training): Executed complex Reinforcement Learning from Human Feedback (RLHF) and Supervised Fine-Tuning (SFT) workflows. Authored detailed SFT Justifications to train models on why a specific response is superior, focusing on logic, syntax, and safety. Conducted Side-by-Side (SbS) evaluations with strict adherence to complex grading rubrics. Performed deep Root Cause Analysis (RCA) to identify model hallucinations and logic failures. 🧮 2. Math Reasoning & Logic (STEM): Generated and evaluated Chain-of-Thought (CoT) prompts for complex mathematical problems (Calculus, Linear Algebra). Verified factual accuracy and logical consistency of model outputs using LaTeX formatting and rigorous proof-checking. Designed "Golden Data" sets (Ground Truth) used to benchmark model performance and train other annotators.

CORE DOMAINS & RESPONSIBILITIES: 🚀 1. Advanced RLHF & SFT (Model Training): Executed complex Reinforcement Learning from Human Feedback (RLHF) and Supervised Fine-Tuning (SFT) workflows. Authored detailed SFT Justifications to train models on why a specific response is superior, focusing on logic, syntax, and safety. Conducted Side-by-Side (SbS) evaluations with strict adherence to complex grading rubrics. Performed deep Root Cause Analysis (RCA) to identify model hallucinations and logic failures. 🧮 2. Math Reasoning & Logic (STEM): Generated and evaluated Chain-of-Thought (CoT) prompts for complex mathematical problems (Calculus, Linear Algebra). Verified factual accuracy and logical consistency of model outputs using LaTeX formatting and rigorous proof-checking. Designed "Golden Data" sets (Ground Truth) used to benchmark model performance and train other annotators.

2024 - 2025

Oracle-Tier AI Trainer & Multidisciplinary Subject Matter Expert

Internal Proprietary ToolingComputer Code ProgrammingClassificationFine Tuning
Code Generation & Engineering: Evaluated code-based prompts in Python, C++, Lua, and SQL. Performed Code Execution & Unit Testing to validate model-generated scripts against edge cases and syntax errors. Refined prompts to improve the model's ability to handle debugging, refactoring, and algorithmic optimization. 🎙️ Audio & Voice Technologies (TTS/ASR): Served as a specialized Voice Talent for Text-to-Speech (TTS) training, providing high-fidelity audio data using professional hardware (192kHz). Evaluated Prosody, Intonation, and Emotional Authenticity in synthesized speech. Worked on cross-lingual tasks (Transcription, Translation, Localization), ensuring Tone Consistency and cultural relevance for UA/RU/PL/EN locales. Key Achievement: Consistently maintained a >98% Quality/Accuracy score across all domains while handling high-priority queues.

Code Generation & Engineering: Evaluated code-based prompts in Python, C++, Lua, and SQL. Performed Code Execution & Unit Testing to validate model-generated scripts against edge cases and syntax errors. Refined prompts to improve the model's ability to handle debugging, refactoring, and algorithmic optimization. 🎙️ Audio & Voice Technologies (TTS/ASR): Served as a specialized Voice Talent for Text-to-Speech (TTS) training, providing high-fidelity audio data using professional hardware (192kHz). Evaluated Prosody, Intonation, and Emotional Authenticity in synthesized speech. Worked on cross-lingual tasks (Transcription, Translation, Localization), ensuring Tone Consistency and cultural relevance for UA/RU/PL/EN locales. Key Achievement: Consistently maintained a >98% Quality/Accuracy score across all domains while handling high-priority queues.

2024

AI Data Specialist & Quality Assurance (Linguistics) / Data annotator

Internal Proprietary ToolingTextEntity Ner ClassificationClassification
Summary: Executed Human-in-the-Loop (HITL) workflows for large-scale Generative AI and ASR (Automatic Speech Recognition) projects. Specialized in high-precision Data Annotation, Data Cleansing, and Linguistic QA to ensure "Ground Truth" accuracy for foundation models across Slavic languages (UA/RU/PL) and English. Key Responsibilities: Multilingual Data Annotation: Performed granular content labeling, Named Entity Recognition (NER), and semantic tagging (Sentiment, Intent) to refine NLP algorithms. Quality Assurance (QA) & Audit: Conducted rigorous peer-reviews on transcribed datasets, maintaining a >98% accuracy rate in capturing prosody, dialects, and edge cases. Dataset Curation: Contributed to the creation of annotated "Golden Sets" for Supervised Learning, validating model outputs against strict guidelines. Operational Efficiency: Managed high-volume workflows involving Data Entry and Transcription, ensuring strict adherence to SLAs and project timelines.

Summary: Executed Human-in-the-Loop (HITL) workflows for large-scale Generative AI and ASR (Automatic Speech Recognition) projects. Specialized in high-precision Data Annotation, Data Cleansing, and Linguistic QA to ensure "Ground Truth" accuracy for foundation models across Slavic languages (UA/RU/PL) and English. Key Responsibilities: Multilingual Data Annotation: Performed granular content labeling, Named Entity Recognition (NER), and semantic tagging (Sentiment, Intent) to refine NLP algorithms. Quality Assurance (QA) & Audit: Conducted rigorous peer-reviews on transcribed datasets, maintaining a >98% accuracy rate in capturing prosody, dialects, and edge cases. Dataset Curation: Contributed to the creation of annotated "Golden Sets" for Supervised Learning, validating model outputs against strict guidelines. Operational Efficiency: Managed high-volume workflows involving Data Entry and Transcription, ensuring strict adherence to SLAs and project timelines.

2023

AI Math Trainer

Internal Proprietary ToolingTextEvaluation Rating
Assessing the factuality and relevance of domain-specific text produced by AI models Crafting and answering questions related to Math Evaluating and ranking domain-specific responses generated by AI models

Assessing the factuality and relevance of domain-specific text produced by AI models Crafting and answering questions related to Math Evaluating and ranking domain-specific responses generated by AI models

2024 - 2025

AI Voice Actor | AI Trainer

Internal Proprietary ToolingAudioEmotion RecognitionAudio Recording
As an AI voice actor at Outlier, I specialise in creating natural-sounding voice recordings of a high quality for AI and machine learning applications. My work supports the development of next-generation voice technologies, for which tone, clarity and emotional authenticity are crucial. I have collaborated with AI/ML teams to enhance training datasets for voice synthesis and natural language applications. I have actively contributed to AI prompt enhancement initiatives, optimising phrasing, tone and delivery to create more engaging, human-like voice outputs. I have received consistent recognition from my Quality Manager for my ability to convey emotion, nuance and realism in my voice acting work, while maintaining a natural and engaging tone. I have adapted to various character roles, scenarios, and emotional ranges, thereby contributing to more diverse and dynamic AI voice libraries.

As an AI voice actor at Outlier, I specialise in creating natural-sounding voice recordings of a high quality for AI and machine learning applications. My work supports the development of next-generation voice technologies, for which tone, clarity and emotional authenticity are crucial. I have collaborated with AI/ML teams to enhance training datasets for voice synthesis and natural language applications. I have actively contributed to AI prompt enhancement initiatives, optimising phrasing, tone and delivery to create more engaging, human-like voice outputs. I have received consistent recognition from my Quality Manager for my ability to convey emotion, nuance and realism in my voice acting work, while maintaining a natural and engaging tone. I have adapted to various character roles, scenarios, and emotional ranges, thereby contributing to more diverse and dynamic AI voice libraries.

2024 - 2025

Education

Z

Zaporizhzhya National University

Master of Science, Applied Mathematics

Master of Science
2009 - 2015

Work History

O

Outlier

AI Voice Actor | AI Trainer

Wroclaw
2024 - Present
O

Outlier

Oracle-Tier AI Trainer & Multidisciplinary Subject Matter Expert

Wroclaw
2024 - Present