智能湖仓研发工程师（上海/深圳）
Core
Building high-performance, governable, and reusable data infrastructure for a unified AI data lake, supporting ad, recommendation, and search business scenarios.
Role type
Senior IC data infrastructure engineer (AI data lake & lakehouse)
Builds
Unified data assets (samples, features, logs, model I/O) and efficient access capabilities for LLM R&D pipelines including data collection, cleaning, feature extraction, embedding generation, and RAG.
Domain
AI/ML, Big Data, Distributed Systems, Data Lakehouse
Deliverable
production ML models
Required skills
Distributed systems design, Storage system internals, Query planning & optimization, Columnar storage formats, Vector search, LLM data pipeline engineering
Preferred skills
Ad/recommendation data platforms, Multi-modal data management, RAG systems, Open-source project contributions (Iceberg, Hudi, Lance, etc.)
Technologies
Spark, Flink, StarRocks, Ray, PyTorch, Java, Scala, C++, Rust, Python, Parquet, Lance, Vortex, Iceberg, Hudi, Paimon, Arrow
Responsibilities
Design and build data infrastructure for storage, metadata, indexing, versioning, and lifecycle management; Optimize end-to-end performance for data writing, scanning, filtering, and training data loading; Manage core data assets for sample construction and feature backtracking; Support LLM R&D linkages from data cleaning to inference backwrite; Construct high-throughput processing and retrieval capabilities for multi-modal and semantic search.
