At Roche you can show up as yourself, embraced for the unique qualities you bring. Our culture encourages personal expression, open dialogue, and genuine connections, where you are valued, accepted and respected for who you are, allowing you to thrive both personally and professionally. This is how we aim to prevent, stop and cure diseases and ensure everyone has access to healthcare today and for generations to come. Join Roche, where every voice matters.The PositionThe Data Scientist - Enterprise Search role is responsible for contributing to the design and development of the next-generation Enterprise Search and information retrieval architectures.This role will build context-aware, generative AI-driven search systems optimized for agentic readiness—enabling autonomous tool orchestration, deep semantic understanding, and intelligent information retrieval at an enterprise scale.The data scientist will develop advanced AI solutions, with a strong focus on Generative AI, LLM-based applications, and scalable data services. This requires to work with large datasets, develop and evaluate machine learning models, and collaborate with cross-functional teams to improve the accuracy, coverage, relevance, and performance of search algorithms at enterprise scale.This role involves direct communication with project stakeholders and contributes to team best practices, while identifying optimization opportunities that enhance the impact of moderately complex data solutions within larger product architectures. You will leverage advanced technical skills to translate business needs into actionable data science initiatives.Job ResponsibilitiesGenerative AI, Agentic AI and LLM OptimizationModel Development & Experimentation: Lead exploratory data analysis, feature engineering, model selection, training, validation, and performance evaluation for machine learning and AI-enabled solutions. Design and evaluate multiple modeling approaches, establish appropriate evaluation metrics, and optimize models for scalability, reliability, and business impact. Experimentation and Innovation: lead experimental projects and drive innovation in enterprise search, exploring novel approaches like GraphRAG or agentic search patterns. Agentic AI Search: Develop and deploy intelligent agentic architectures that can interact with and enhance the enterprise search experienceRAG Experimentations (RAG Evaluation Framework): Design and conduct Retrieval-Augmented Generation experiments to evaluate and improve search relevance and performanceLLM Model Evaluation: Evaluate the performance of Large Language Models in various enterprise search contexts, ensuring they meet business requirements and performance standardsAdvanced Prompt Engineering: Design and optimize prompts to programmatically enhance the interaction and effectiveness of search queries and responsesBusiness Problem Solving & Decision Support Partner closely with business stakeholders and product teams to translate complex business challenges into analytical approaches, ML solutions, and scalable intelligence capabilities. Support data-driven prioritization and strategic decision making through actionable insights, predictive models, and operational intelligence. Conducts A/B testing and experiments to assess the performance of search models and algorithmsConsultancy Provide expert consultancy on data science and machine learning best practices, guiding internal teams and stakeholdersPoC and knowledge sharing: Design and lead proof-of-value (PoV) projects, conduct knowledge-sharing sessions, and deliver impactful demos to showcase capabilities and gather feedbackData Engineering & Vector DatabasesData Engineering & Processing: Work with structured and unstructured data, building efficient pipelines for data ingestion, preprocessing, and feature engineeringVector database experimentation: Conduct experiments with vector databases to improve the efficiency and accuracy of our search systemsManaging embeddings: Implement and optimize techniques for embeddings generation, indexation and retrieval to support advanced search queries and retrieval capabilitiesRetrieval & relevance tuning: Develop, evaluate, and tune retrieval algorithms to optimize search precision, recall, and relevance for diverse datasetsData quality: Design data enhancement modules that extract, enrich, and validate document content and metadata, directly improving downstream model context, search recall, and agentic reasoningModel Lifecycle and integrationDeployment, Testing, and Training of ML Models and Endpoints: Develop, deploy, and continuously refine machine learning models and endpoints to enhance search functionalities. Conduct rigorous testing and validation to ensure model accuracy and reliabilityDevelopment of ML Models for search: Create and deploy advanced runnable models, such as entity extraction models and metadata augmentation, to expand the capabilities of our search solutionsMLOps & Monitoring: Implement best practices for model deployment, versioning, monitoring, and performance optimizationAPI & MCP Interoperability: Interface with enterprise search engines, platforms, and other APIs to enhance our search functionalities and integrations. Familiarity with MCP and protocols for agent interoperabilityQualificationsEducation / ExperienceMaster’s degree or PhD in Data Science, Computer Science, Statistics, Mathematics, Engineering, Artificial Intelligence, or a related quantitative field. Demonstrated experience as a rising expert developing predictive models and leading specific analytical modules or project componentsProven track record of taking full accountability for the quality and timely delivery of analytical tasks and troubleshooting complex data issues independentlyExperience working effectively on moderately complex data science problems and understanding how contributions fit into medium-sized data architecturesTechnical SkillsShows strong proficiency in programming languages, particularly PythonHas proven experience as a Data Scientist, preferably with a focus on information retrieval and NLPHas a solid understanding of natural language processing (NLP) techniques and toolsPossesses hands-on experience with machine learning and deep learning frameworks and libraries (e.g., TensorFlow, PyTorch, scikit-learn)Familiarity with cloud platforms and services, particularly AWS or Microsoft AzureFamiliarity with version control systems (e.g., Git) and agile development practicesPast experience with search engines and technologies (e.g. Elasticsearch, Solr, or Lucene) and solid understanding of search algorithms, information retrieval, and relevancy tuning is a plusDemonstrates excellent analytical and problem-solving skills, with the ability to work with large, complex datasetsProven ability to translate well-defined business questions into clear analytical problems and technical solutionsAdditional QualificationsStrong communication and collaboration skills, with the ability to manage direct communication with immediate project stakeholdersAbility to actively integrate feedback from technical peers and junior team membersProactive mindset to identify potential optimizations or new analytical approaches within the project scopeAbility to work autonomously to achieve goals and deliver results, while actively collaborating with team members to meet shared team objectives. Experience in healthcare, pharmaceutical, or other regulated industries is a plusThis vacancy will remain open until August 20To ensure a fair selection process, applications will be reviewed and candidates will be shortlisted only after that dateWho we areA healthier future drives us to innovate. Together, more than 100’000 employees across the globe are dedicated to advance science, ensuring everyone has access to healthcare today and for generations to come. Our efforts result in more than 26 million people treated with our medicines and over 30 billion tests conducted usin