Dinesh K.

Key Skills

Software

Other

Top Subject Matter

Medical domain expert evaluation for RLHF and AI model training (pharmaceutical/clinical content)

Medical affairs domain—AI output evaluation and ranking

Medical domain red-teaming and evaluation of AI outputs

Top Data Types

Text

Document

Top Task Types

RLHF

Red Teaming

Prompt + Response Writing (SFT)

Freelancer Overview

Global Medical Director – Rare Liver Disease | Ipsen (Consultancy) | AI/RLHF medical content evaluation, ranking, correc. Brings 22+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Other. Education includes Master of Business Administration, London Business School (2017) and Postgraduate Certificate in Pharmaceutical Medicine, Cardiff University (2010). AI-training focus includes data types such as Text and labeling workflows including RLHF, Evaluation, and Rating.

Labeling Experience

Global Medical Director – Rare Liver Disease | Ipsen (Consultancy) | AI/RLHF medical content evaluation, ranking, correction, and red-teaming

OtherTextRLHF

Provided medical domain expertise to evaluate and improve AI-generated medical content for RLHF and AI model training, emphasizing clinical accuracy, safety, and guideline compliance. Performed critical review and response correction workflows by identifying hallucinations, logical gaps, and inconsistencies in scientific and clinical outputs. Designed and evaluated prompt and response quality using structured medical evaluation rubrics to support model training objectives. • Conducted clinical accuracy reviews across rare liver disease and broader medical domains. • Performed response correction and scientifically nuanced rewriting of incorrect or incomplete AI outputs. • Applied red-teaming/adversarial testing approaches using complex clinical scenarios and edge-case queries. • Reviewed compliance with EMA, MHRA, NICE, FDA, and promotional/regulatory standards for AI outputs.

2025 - Present

Medical Affairs Consultant – Rare Disease & Oncology | Astellas Pharma | AI output evaluation, ranking, and correction for medical content

OtherText

Evaluated AI-generated scientific narratives and medical communications for accuracy, balance, completeness, and logical flow as part of AI model training support. Performed pairwise response ranking and comparative assessment to identify safer and more guideline-consistent outputs. Corrected and refined AI responses to ensure clinically appropriate nuance suitable for medical affairs usage. • Ranked and compared AI outputs for factual correctness and evidentiary alignment. • Rewrote or corrected AI-generated content to close knowledge gaps and improve clarity. • Assessed safety concerns and mitigations for potentially misleading medical guidance. • Ensured therapeutic-area-specific consistency with treatment guidelines and standards.

2024 - 2025

International Medical Director (RNAi – 40+ countries) | Alnylam Pharmaceuticals | AI red-teaming/evaluation support for medical content

OtherTextRed Teaming

Used clinical and regulatory expertise to probe AI models with complex medical scenarios to identify unsafe, misleading, or hallucinatory responses. Conducted systematic evaluation across multiple therapeutic areas to strengthen model reliability for medical communications. Supported AI training workflows by applying guideline, mechanistic, and safety checks informed by established regulatory frameworks. • Performed adversarial/red-teaming style queries using clinical edge cases and ambiguity. • Assessed mechanistic accuracy and clinical safety across oncology, immunology, and cardiovascular contexts. • Evaluated outputs against regulatory and promotional standards (EMA/MHRA/NICE/FDA frameworks). • Identified guideline inconsistencies and logical gaps to improve training signal quality.

2021 - 2024

Medical Director – Immunology | Sanofi Genzyme (UK & Ireland) | Prompting and supervised correction of AI responses for training-quality medical content

OtherTextPrompt + Response Writing (SFT)

Developed and refined prompt strategies and structured evaluation rubrics for medical literature synthesis and AI-assisted medical communications. Reviewed and corrected AI-generated content for factual correctness and guideline adherence, improving training-quality outputs. Applied expert medical writing judgment to ensure responses were precise and appropriately nuanced for clinical/scientific contexts. • Created prompt approaches for medical content development and evaluation. • Conducted response correction and scientifically accurate rewriting. • Checked for hallucinations, evidence gaps, and guideline mismatches. • Used structured assessment criteria to support repeatable quality control.

2019 - 2021

Education

Work History