Software Engineer, Memory & Search
$320k - $485k • Remote • New York City, NY; New York City, NY | Seattle, WA; Remote-Friendly, United States; San Francisco, CA | New York City, NY | Seattle, WA
Posted 7h ago
Job Location
New York City, NY; New York City, NY | Seattle, WA; Remote-Friendly, United States; San Francisco, CA | New York City, NY | Seattle, WA
Tech Stack
Remote Work Policy
Fully remote
Categories
AI Agent Engineer
About the job
Anthropic is seeking Software Engineers to join the Memory & Search team, responsible for how Claude remembers users and finds information across Claude.ai, Cowork, and Claude Code. This role involves both user-facing experiences, such as how memory is displayed and edited, and the underlying infrastructure for memory services, indexing, retrieval, and vector search. The team has recently transitioned memory to a dynamic, editable system unified across products, and improved search functionality significantly. The focus is now on developing compelling product features that leverage memory and search, requiring engineers with experience in shipping user-facing products at scale, a strong sense for product quality, and expertise in building agentic systems where models, tools, and user interfaces collaborate. You will drive product experiments, analyze usage, and contribute to the team's product-centric approach. This role also involves close collaboration with product, design, privacy, legal, and safeguards teams to ensure features meet privacy and enterprise requirements.
Responsibilities
- Design, build, and ship memory and search experiences for Claude.ai, Cowork, and Claude Code, covering backend services, model tools, and user interfaces.
- Conduct product experiments on the memory and search stack, including hypothesis formation, demo building, user testing, measurement, and decision-making.
- Shape how Claude utilizes memory and retrieval within its agentic loop, influencing model storage, search triggers, result ranking, presentation, and user control.
- Define and track metrics for memory's usefulness and build evaluations to maintain system quality.
- Enhance search relevance and recall across conversations and memories by working with embeddings, hybrid retrieval, ranking, and result presentation interfaces.
- Collaborate with product, design, privacy, legal, and safeguards teams to ship features that comply with user consent and enterprise requirements.
- Improve product quality through design and code reviews, setting standards for latency, polish, and reliability.
- Participate in on-call rotations for owned systems.
Requirements
- Experience building and operating user-facing products with large user bases, including their backend services.
- Experience building agentic systems or LLM-powered products, including tool use and retrieval.