AI-Native Database Architect & Researcher (Vector/LLM)

ByteDance; San Jose (CA); On-site; USD 156,000 - 387,600; Full time

Job summary

ByteDance is seeking a Software Engineer/Researcher to build AI-native database systems in San Jose. You will architect databases that fuse structured, unstructured and vector data, design fast storage engines and scalable vector search, and develop LLM-driven query processors.

The role involves RAG infrastructure work and contributions to open-source communities. The team values strong systems or AI research background and collaboration across research, engineering and product teams to

Qualifications

• Experience in core database systems or ML systems.
• Strong system-level coding in C++, Rust or Go.
• Ability to translate ideas into production systems across teams.

Responsibilities

• Architect AI-native databases integrating structured, unstructured, and vector data.
• Design storage engines for embedding ingestion and multimodal retrieval.
• Build scalable vector search systems with low latency.
• Develop AI-augmented query processors using LLMs for semantic parsing.
• Collaborate on RAG infrastructure and memory backends.
• Advance learned index structures and AI-integrated transactions.
• Publish and contribute to research and open-source communities.

Skills

C++

Rust

Go

Distributed systems

Storage engine design

Vector retrieval/ANN

AI infra/LLMs

Cross-team collaboration

Education

Bachelor’s/Master’s/PhD in CS or related field

Tools

Faiss

Milvus

DuckDB

ClickHouse

TiKV

RocksDB

Job description

ByteDance is seeking a Software Engineer/Researcher to build AI-native database systems in San Jose. You will architect databases that fuse structured, unstructured and vector data, design fast storage engines and scalable vector search, and develop LLM-driven query processors.

The role involves RAG infrastructure work and contributions to open-source communities. The team values strong systems or AI research background and collaboration across research, engineering and product teams to