Senior AI Engineer - Search & Recommendation - Patient Team (x/f/m)
About this role
The Patient team at Doctolib builds search and recommendation engines for a health companion that helps millions of patients across Europe navigate medical care. Architectural ownership covers full retrieval, ranking, and reranking pipelines that deliver trustworthy and fast insights for web and mobile interfaces.
Daily work involves designing production search architectures, setting up evaluation frameworks, and establishing event feedback loops before building custom machine learning models. Collaboration happens daily with product managers and software engineers to ship AI-driven features focused on latency, reliability, and measurable user value.
Ideal candidates bring significant background as software or AI engineers with hands-on experience deploying retrieval systems to production environments. A methodical, architecture-first mindset that prioritizes robust baselines, metrics like nDCG and Recall@k, and rigorous offline and online evaluation fits well within the team.
What you will do
- Design and build production search and recommendation architectures using vector search, semantic retrieval, and managed rerankers.
- Establish solid baselines through prompts and RAG before implementing custom machine learning solutions.
- Integrate offline and online evaluation alongside observability at every stage of the retrieval pipeline.
- Set up data and event feedback loops that drive iteration and feed deeper machine learning models.
- Improve search relevance and ranking on patient-facing products to elevate result quality.
- Manage production quality metrics including latency reliability, monitoring, and long-term maintainability.
- Partner with machine learning engineers, product managers, and software engineers to ship impactful AI features.
What you need
- Demonstrated production deployment experience shipping algorithms to ECS-based services on AWS.
- Significant background as a software or AI engineer with experience deploying search or recommendation systems.
- Hands-on practice building end-to-end retrieval, ranking, and reranking pipelines familiar with metrics like nDCG, MAP, and Recall@k.
- Proficiency in turning foundation models and off-the-shelf components into production systems such as embeddings and vector search.
- An architecture-first approach prioritizing standard tooling, strong baselines, and evaluation loops before custom machine learning.
- Experience implementing evaluation and observability across retrieval, ranker, and reranker stages including A/B testing and bias handling.
Nice to have
- Experience at a B2C marketplace such as e-commerce, hospitality, or travel
- Additional methodologies in pattern mining, recommendation systems, experimentation, or causal inference
- Familiarity with search engines or information retrieval concepts
- Exposure to learning-to-rank or feature engineering alongside machine learning engineers
- Background working in healthcare or other regulated domains under GDPR or HDS
Benefits
- Free comprehensive health insurance for you and your children
- 25 days of paid vacation per year, plus up to 14 days of RTT
- Free mental health and coaching services through Moka.care
- Work from abroad for up to 10 days per year via flexibility days
- Lunch vouchers worth €8.50 per working day with partial company coverage
- 50% reimbursement of public transport subscription
Skills and tech stack
Questions candidates ask
What is the tech stack used for building these AI features?
Solutions are built on a single, fully cloud-native platform supporting web and mobile interfaces. The broader stack includes Rails, TypeScript, Java, Python, Kotlin, Swift, and React Native.
What does the interview process look like for this position?
The hiring process starts with an initial interview with a member of the Talent Acquisition team. Subsequent steps include a coding interview, an agentic system design session, a behavioral interview, and a reference check.
What kind of evaluation and monitoring is expected in the role?
Candidates must build evaluation and observability into every stage of the pipeline, covering both offline and online evaluation. This includes A/B testing, position-bias handling, and production monitoring for latency and reliability.
What is the primary focus of the Patient team?
The Patient domain ensures everyone has better access to necessary care and active prevention tools. The team designs search and recommendation engines for a health companion helping millions of patients navigate to exact care.