AI 原生多模数据库内核研发工程师
Core
Design and develop the kernel architecture for a multi-modal database engine handling high-concurrency storage of vectors, text, and files, and optimizing fusion queries for AI workloads.
Role type
Senior IC database kernel engineer (multi-modal storage & AI fusion)
Builds
Multi-modal storage engine, unified smart routing tables, and distributed query execution paths for real-time retrieval.
Domain
Database systems, distributed storage, AI infrastructure
Deliverable
production ML models
Required skills
C/C++ or Rust, Linux kernel programming, multi-threading and concurrency control, memory management, SIMD/AVX512 hardware acceleration, vector index algorithms (HNSW/IVF), distributed computing frameworks (Ray, Spark, Flink)
Preferred skills
OpenSearch/Elasticsearch/Lucene source code expertise, Milvus/FAISS implementation knowledge, community contributions to big data ecosystems
Technologies
Ray, Spark, Flink, Iceberg, ClickHouse, HNSW, IVF, SIMD, AVX512
Responsibilities
Design multi-modal storage engine architecture and persistent storage mechanisms; Implement cross-modal fusion query execution engines and optimize complex operators; Optimize memory management and caching strategies for AI workloads; Solve engineering challenges in index real-time updates and tail latency reduction.
Seniority
Senior, hands-on IC