AI Data Engineer (Materials Data Extraction)

About the role

Patsnap is looking for a Data Engineer or Senior Data Engineer to build and evolve the data foundation behind its global innovation intelligence and AI-powered products. The role works with large-scale patent and related datasets, turning complex raw information into accurate, reliable and accessible data. It is a hands-on engineering position with ownership across the full data lifecycle, from ingestion and parsing to storage, quality, performance and availability. The position is in-office in the Singapore office.

Key responsibilities

• Process and extract structured information from materials-related patents, scientific literature and technical documents.
• Design and test prompts, input formats and extraction workflows across different models and datasets.
• Build LLM workflows covering input preparation, model calls, prompt versioning, validation, evaluation, retries and human review.
• Analyse model failures and improve results by changing prompts, rules, data or workflow design.
• Help build gold-standard datasets, annotation guidelines, quality checks and automated evaluations.
• Use Python and SQL for data cleaning, transformation, batch processing and result analysis.
• Improve the accuracy, reliability, speed and cost efficiency of extraction workflows, and turn successful experiments into reusable scripts, tools or automated workflows.

Required profile

• Bachelor's degree in Computer Science, Software Engineering, Data Science, Artificial Intelligence, Materials Science or a related discipline.
• Interest in LLM applications and experience with at least one large language model or AI coding tool.
• Strong logical thinking and a structured approach to experimentation, tracking changes and evaluating their impact.
• Attention to data quality and good communication and teamwork skills, with the ability to read English technical documentation.

Required skills

• Good programming fundamentals, using Python or Java for basic development and data processing.
• Basic SQL and experience with structured data.
• Nice to have: LLM, prompt engineering, RAG, agents, information extraction or automated evaluation experience; data processing, ETL, web crawling, data annotation or knowledge bases; model APIs such as OpenAI, Gemini or Claude; machine learning, NLP, named entity recognition or structured information extraction; materials science, patents or research data background; AI coding tools such as Cursor, Codex or Claude Code.