SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 320,000 - 485,000 / annual
Anthropic is seeking a Staff+ Software Engineer to own the vector database infrastructure that powers Claude's retrieval capabilities. This is a foundational role within a small, core team where architectural decisions remain open and early joiners will shape the system's evolution.
You will be responsible for the vector database layer serving Claude's retrieval workloads across web search (claude.ai, API, agentic surfaces), conversation memory, and reinforcement learning training pipelines. The role sits within Anthropic's Databases organization and collaborates closely with the Search Platform team (web search indexing and serving) and Memory & Search team (claude.ai search and memory features).
Key responsibilities include:
- Owning the core vector database layer: index structures, storage layout, query execution, and serving infrastructure
- Ensuring production readiness through benchmarking, capacity planning, workload onboarding, and operational tooling at scale
- Driving recall, latency, and cost trade-offs via quantization, filtered search, index build/refresh strategies, and compute/storage separation
- Scaling throughput by an order of magnitude as research and training demands grow; identifying architectural, hardware, or engineering bottlenecks
- Extending the retrieval stack toward hybrid retrieval (lexical search, fused ranking alongside vector search)
- Setting evaluation methodology: recall targets, latency objectives, cost per query, and regression detection
- Setting technical direction for a growing team through design reviews, architectural decisions, and mentoring
- Sharing on-call responsibilities for owned systems
You will be hands-on in the code while also deciding how system pieces fit together. The team operates at scale with real users and production on-call demands.
REQUIREMENTS:
- Direct experience building a vector database or approximate nearest-neighbor search system in production at scale
- Deep knowledge of ANN index structures (graph-based, clustering-based, or disk-resident approaches), vector quantization, and filtered search
- Experience with search relevance and ranking
- Strong systems programming skills in Rust, C++, or Go, including performance-critical and memory-intensive code
- Experience designing and operating large-scale distributed storage or serving systems with strict latency and reliability requirements
- Track record of independently scoping and delivering ambiguous, multi-quarter technical projects
- Strong written and verbal communication with experience driving alignment across teams with different priorities
- Interest in AI safety and Anthropic's mission
PREFERRED QUALIFICATIONS:
- Experience with object-storage-backed or disaggregated compute/storage database designs
- Experience building or operating full-text search engines (inverted indexes, BM25) or hybrid retrieval systems
- Familiarity with embedding models and retrieval demands of LLM training and inference
- Experience using GPUs or other accelerators for index build or search
- Prior experience as technical lead on a database or search infrastructure team
Minimum education: Bachelor's degree or equivalent combination of education, training, and professional experience in a field relevant to the role. Hybrid policy: at least 25% in-office time, though some roles may require more. Visa sponsorship available.
