Research Staff

Role overview Conduct applied research to improve information retrieval for systems that need to find useful context across different formats. The work spans retrieval methods, multimodal models, and search pipelines augmented by large language models, with opportunities to influence both product performance and the research agenda.

Responsibilities - Lead or contribute to applied information retrieval research projects. - Develop and evaluate approaches to fine-grained late interaction and multimodal retrieval. - Investigate how large language models can generate, rank, and use search results. - Improve retrieval performance and efficiency through research that informs the search platform. - Build evaluation methods for realistic, agent-oriented use cases and improve retrieval data signals. - Collaborate with engineering and product teams, and share findings through papers, articles, or conferences.

Requirements - Experience in information retrieval, embeddings, or a related field. - An advanced degree in a relevant area or equivalent experience with impactful research projects. - Python proficiency, including the ability to implement models from scratch. - Broad understanding of machine learning, including algorithms, training, data, and efficiency. - Strong written and verbal communication skills. - Interest in advancing retrieval beyond single-vector similarity approaches.

Nice to have - Experience with transformers, interaction models, or vector search systems. - Experience designing and running large-scale training experiments.

Benefits and work setup - Full-time role based in San Francisco or remote. - Compensation includes equity; health, dental, and vision coverage are offered. - Visa sponsorship and relocation support are available. - Other listed support includes professional development, conference attendance, transportation, wellness, and food.