Biomedical Data Enrichment Expert

General information - Duration: 3 months, possibly extended with 3 more months - No. of working hours: 24 - 36 hours per week - Location: 1 day per week onsite in Amsterdam - Contract type: B2B contract - VISA sponsorship: No - Location: You must live in NL at the time of application What is the project about? You will be creating a high-quality “master answer key” or benchmark dataset. You take real, everyday conversational questions asked by doctors or researchers (natural-language queries) and translate a vague medical question into a highly structured clinical framework. You then match them with the perfect, manually verified search results from biomedical databases (Embase). The engineering team uses this “gold standard” to test the AI. If the AI changes its algorithm, they run it against your answer key to ensure its performance didn’t degrade. You read the AI’s summary of a medical topic and ensure it matches the actual scientific data in the cited studies. You evaluate it using strict metrics like scientific accuracy, completeness, clarity, and citation truthfulness. You look at the data from your evaluations (e.g., “The AI is failing to understand pediatric dosages 30% of the time” ). You then explain this finding to software engineers and product managers in a way they understand. Responsibilities - Build and maintain gold-standard evaluation sets for natural-language queries to biomedical databases. - Define biomedical search use cases and develop systematic review strategies using frameworks such as PICO and SPIDER. - Evaluate generated queries for syntax, scientific accuracy, controlled vocabulary, field codes, and Boolean logic. - Assess AI-generated summaries and supporting references using defined quality and scientific evaluation metrics. - Analyze …

Do you have experience in the biomedical domain and do you have some AI / API skills?