About the role: We are looking for Research Engineersto help us redesign how Claude interacts with external datasources. Many of the paradigms for how data and knowledge bases areorganized assume human consumers and constraints. This is no longertrue in a world of LLMs! Your job will be to design newarchitectures for how information is organized, and train languagemodels to optimally use those architectures. Responsibilities: *Designing and implementing from scratch new informationarchitecture strategies * Performing finetuning and reinforcementlearning to teach language models how to interact with newinformation architectures * Building “hard” knowledge base evalsets to help identify failure modes of how language models workwith external data * Designing and evaluating advanced agenticsearch capabilities. You may be a good fit if you: * Are a veryexperienced Python programmer who can quickly produce reliable,high quality code that your teammates love using * Have goodmachine learning research experience * Have experience developingsoftware that utilizes Large Language Models such as Claude * Areresults-oriented, with a bias towards flexibility and impact * Pickup slack, even if it goes outside your job description * Enjoy pairprogramming (we love to pair!) * Want to partner with world-classML researchers to develop new LLM capabilities * Care about thesocietal impacts of your work * Have clear written and verbalcommunication Strong candidates will also have experience with: *Collaborating with product teams to quickly prototype and deliverinnovative solutions * Building complex agentic systems thatutilize LLMs * Developing scalable distributed informationretrieval systems, such as search engines, knowledge graphs, RAG,indexing, ranking, query understanding, and distributed dataprocessing J-18808-Ljbffr