Data Science Lead (AI)

TL;DR

Data Science Lead (AI/LLM): Leading the machine learning and GenAI architecture for a production-grade pharmaceutical compliance platform with an accent on LLM-driven retrieval (RAG) and regulatory validation. Focus on designing robust evaluation frameworks, implementing explainability mechanisms, and scaling AI workflows to an enterprise level.

Location: Remote

Company

hirify.global provides high-impact AI expertise for strategic enterprise initiatives in regulated sectors.

What you will do

• Design and evolve AI/ML architecture for compliance validation workflows in a pharmaceutical environment.
• Optimize LLM-driven retrieval and validation pipelines using RAG-based systems.
• Establish evaluation frameworks, benchmarking strategies, and continuous improvement loops.
• Implement explainability and traceability mechanisms to meet strict regulatory standards.
• Collaborate with ML Engineers and Backend teams to productionize AI components.
• Drive technical decisions regarding embeddings, vector databases, and retrieval strategies.

Requirements

• Strong hands-on background in Data Science and applied Machine Learning.
• Proven experience designing and deploying production-grade LLM-based systems.
• Practical expertise with Retrieval Augmented Generation (RAG) and vector databases.
• Proficiency in Python and modern AI frameworks.
• Experience designing model evaluation and validation frameworks.
• Ability to operate effectively within strictly regulated or compliance-heavy environments.

Nice to have

• Experience in pharmaceutical, healthcare, or other regulated industries.
• Exposure to explainable AI (XAI) methodologies or frameworks.
• Expertise in document intelligence and NLP-heavy pipelines.
• Experience building enterprise-scale AI platforms rather than PoCs.

Culture & Benefits

• Competitive salary.
• Opportunity to work in a multinational environment on international projects.
• Comprehensive healthcare coverage.
• Stable, long-term B2B contract.
• Fully remote working model.