I’m Pavlo Dragan, a Ph.D.-level political scientist and linguist based in Kyiv. With nearly two decades of experience in high-stakes text analysis, translation quality work, and bilingual (Ukrainian/Russian) linguistic QA, I’m focused on building safer, more reliable AI systems—especially in areas tied to RLHF/SFT evaluation. Today I evaluate and stress-test generative AI outputs on geopolitical topics, checking for factual accuracy, hallucinations, bias, and logical coherence in multi-turn responses. I also support alignment-focused workflows through prompt engineering, adversarial red teaming, and structured fact-checking, while keeping strong standards for clarity, terminological consistency, and cultural nuance.…

Pavlo Dragan

I’m Pavlo Dragan, a Ph.D.-level political scientist and linguist based in Kyiv. With nearly two decades of experience in high-stakes text analysis, translation quality work, and bilingual (Ukrainian/Russian) linguistic QA, I’m focused on building safer, more reliable AI systems—especially in areas tied to RLHF/SFT evaluation. Today I evaluate and stress-test generative AI outputs on geopolitical topics, checking for factual accuracy, hallucinations, bias, and logical coherence in multi-turn responses. I also support alignment-focused workflows through prompt engineering, adversarial red teaming, and structured fact-checking, while keeping strong standards for clarity, terminological consistency, and cultural nuance.…

Available to hire

I’m Pavlo Dragan, a Ph.D.-level political scientist and linguist based in Kyiv. With nearly two decades of experience in high-stakes text analysis, translation quality work, and bilingual (Ukrainian/Russian) linguistic QA, I’m focused on building safer, more reliable AI systems—especially in areas tied to RLHF/SFT evaluation.

Today I evaluate and stress-test generative AI outputs on geopolitical topics, checking for factual accuracy, hallucinations, bias, and logical coherence in multi-turn responses. I also support alignment-focused workflows through prompt engineering, adversarial red teaming, and structured fact-checking, while keeping strong standards for clarity, terminological consistency, and cultural nuance.

See more

Skills

Experience Level

Language

Ukrainian
Fluent
Russian
Fluent
English
Fluent

Work Experience

AI Data Specialist & Domain Expert (Political Science) (RLHF & Model Evaluation, Adversarial Red Teaming, Fact-Checking) at Confidential AI Platforms (via Data Annotation / Upwork)
January 1, 2023 - Present
Evaluate and rank complex, multi-turn AI responses on geopolitical topics using HHH (helpfulness, honesty, harmlessness). Perform rigorous hallucination detection via fact-checking of historical events and political figures. Design constraint-based adversarial tests to identify safety vulnerabilities related to political disinformation or hate speech. Assess logical coherence and ensure culturally appropriate reasoning across Ukrainian and Russian contexts.
AI Data Specialist & Domain Expert (Political Science) at Confidential AI Platforms (via Data Annotation / Upwork)
January 1, 2023 - Present
RLHF & Model Evaluation: Evaluate and rank complex, multi-turn AI responses on geopolitical topics (e.g., international treaties, conflict resolution) based on helpfulness, honesty, and harmlessness (HHH). Hallucination Detection: Rigorous fact-checking of AI-generated content regarding historical events and political figures to identify subtle factual errors and fabrications. Adversarial Red Teaming: Design constraint-based prompts to stress-test model safety guidelines, specifically looking for vulnerabilities in generating political disinformation or hate speech. Reasoning Evaluation: Assess the logical coherence of AI arguments in Ukrainian and Russian, ensuring the model captures cultural nuances and avoids US-centric bias in local contexts.
Senior Freelance Translator & Linguistic QA at Upwork / Direct Clients
January 1, 2018 - Present
Localization Expert: Successfully delivered over 3,200 projects involving the translation and cultural adaptation of legal, technical, and business documentation for clients like Google, Hitachi, and Johnson & Johnson. Quality Assurance: Performed linguistic QA for high-volume projects, ensuring 100% adherence to terminological consistency and style guides. Subject Matter Expertise: Applied specialized knowledge of contract law and international standards to translate complex agreements, ensuring zero semantic loss.
Project Manager & Translator at Participatory Culture Foundation (Amara)
January 1, 2014 - Present
Managed distributed teams for large-scale subtitling and localization projects. Oversaw quality control processes to ensure linguistic accuracy across multiple Slavic languages.

Education

Ph.D. (Candidate of Sciences) in Political Science at Taras Shevchenko National University of Kyiv
January 1, 2005 - January 1, 2009
Master of Arts in International Relations & Translation Studies at Taras Shevchenko National University of Kyiv
January 1, 2003 - January 1, 2005
Bachelor of Arts, International Relations and National Security Studies at Taras Shevchenko National University of Kyiv
January 1, 1999 - January 1, 2003
Ph.D. (Candidate of Sciences) in Political Science at Taras Shevchenko National University of Kyiv
January 1, 2005 - January 1, 2009
Master of Arts in International Relations & Translation Studies at Taras Shevchenko National University of Kyiv
January 1, 2003 - January 1, 2005
Bachelor of Arts, International Relations and National Security Studies at Taras Shevchenko National University of Kyiv
January 1, 1999 - January 1, 2003

Qualifications

Google AI Professional Certificate
September 1, 2026 - September 28, 2026

Industry Experience

Education, Government, Media & Entertainment, Non-Profit Organization, Professional Services, Software & Internet

Skills

Experience Level