Software Engineer, Memory & Search

Anthropic•New York City, NY

4d•Hybrid

About The Position

Anthropic is hiring Software Engineers for its Memory & Search team, which builds how Claude remembers users and finds information across Claude.ai, Cowork, and Claude Code. The work spans user-facing memory and search experiences as well as the underlying memory service, indexing, retrieval, and vector search infrastructure. The platform has moved to a dynamic, editable, on-demand memory system shared across products, and the team is now exploring product experiences that make memory a compelling part of using Claude. Engineers will run short, demo-driven experiments, evaluate them with real usage, and help the team work with a strong product focus. The role calls for experience shipping user-facing products at scale and building agentic systems in which models, tools, and interfaces work together. Memory is a trust-sensitive area involving privacy, consent, and enterprise controls. The engineer will collaborate with product, design, privacy, legal, and safeguards partners, treating these considerations as part of product design. The position shares on-call responsibility for the systems owned by the team. Anthropic’s mission is to create reliable, interpretable, and steerable AI systems that are safe and beneficial for users and society. The company is a public benefit corporation headquartered in San Francisco. Its current hybrid policy expects staff to work from an office at least 25% of the time, though some roles may require more. The company sponsors visas, subject to role and candidate circumstances. Compensation is an annual salary of $320,000–$485,000 USD. The stated minimum education is a bachelor’s degree or an equivalent combination of education, training, and/or experience, in a field relevant to the role as demonstrated through coursework, training, or professional experience. Required years of experience correlate with the internal job level.

Requirements

Experience building and operating user-facing products used by many people, including their backend services.
Experience building agentic systems or LLM-powered products, including tool use, retrieval, and prompt or context design.
Strong programming skills in Python and TypeScript, with experience in backend services and web front ends.
Track record of taking a product or feature from idea to launch and iterating based on usage data and user feedback.
Experience defining product metrics and running experiments to inform decisions.
Strong product judgment and ability to explain why design choices matter to users.
Strong written and verbal communication, including experience working closely with product managers, designers, and non-engineering partners.
Interest in AI safety and Anthropic’s mission.
Bachelor’s degree or an equivalent combination of education, training, and/or experience; relevant field of study demonstrated through coursework, training, or professional experience. Years of experience required vary with internal job level.

Nice To Haves

Experience as a founder, early-stage engineer, or in taking a product from zero to one.
Experience building search or retrieval systems, including embeddings, vector search, hybrid lexical and semantic retrieval, or ranking.
Experience building evaluations for LLM or ML systems, or measuring retrieval and ranking quality.
Experience with consumer products with strong retention or personalization systems.
Experience shipping features that handle sensitive data or meet enterprise requirements such as customer-managed encryption keys, data residency, or HIPAA.
Experience with latency-sensitive, performance-critical product work.
Experience mentoring engineers or setting technical direction for a team.

Responsibilities

Design, build, and ship memory and search experiences in Claude.ai, Cowork, and Claude Code, owning them end to end from backend services through model tools to user interfaces.
Run product experiments on the memory and search stack by forming hypotheses, building demos, testing them with users, measuring results, and deciding whether to persevere or pivot.
Shape how Claude uses memory and retrieval in its agentic loop, including what it stores, when it searches, how results are ranked and presented, and how users retain control.
Define and track product metrics for memory usefulness, and build evaluations that maintain quality as the system evolves.
Improve search relevance and recall across conversations and memories, working with embeddings, hybrid retrieval, ranking, and result presentation.
Partner with product, design, privacy, legal, and safeguards teams to ship features that respect user consent and enterprise requirements, including customer-managed encryption keys and data-retention policies.
Raise product quality through design reviews, code reviews, and standards for latency, polish, and reliability.
Share on-call responsibility for the systems owned.