Data Engineer, Global Risk & Anti-Fraud

工作职责

1. 负责账号安全、反作弊、营销风控、内容安全等核心风控业务的数据基建,构建风控指标体系,负责离线/实时风控特征库(Feature Store)的设计与开发。

2. 负责高并发、低延迟的实时风控数据链路建设,为风控引擎和算法模型提供高质量数据支持,包括黑样本库建设、设备指纹数据加工、用户行为序列特征提取等。

3. 负责端云安全数据(如埋点、设备环境、网络特征等)的采集规范制定、质量监控与规范性治理,保障风控底层数据源的准确性、完整性与时效性。

4. 负责风控数据资产的沉淀与复用,保障生产数据的质量与 SLA;跨部门协同风控策略、算法及业务团队,快速响应黑灰产对抗过程中的数据需求。

任职要求

1. 本科及以上学历,计算机相关专业,1年及以上大数据开发经验,具备丰富的实时或离线数据体系建设经验。

2. 扎实的 Java/Scala/Python 编程基础,精通 SQL,具备海量数据开发、复杂逻辑处理及性能调优能力。

3. 熟悉 Hadoop/Spark/Hive 等离线大数据生态,精通 Flink 等实时计算框架,有流批一体或高吞吐、低延迟实时流处理落地经验。

4. 熟悉 ClickHouse/Doris 等 OLAP 引擎,掌握数据仓库建模理论(如维度建模),具备优秀的数据抽象和架构设计能力。

5. 具备强烈的责任心和业务 Sense,能够深入理解黑灰产作弊逻辑,从数据视角主动发现业务风险点。

【加分项】

1. 有业务安全、风控、反作弊数仓开发经验,熟悉设备指纹、账号安全体系或黑产对抗逻辑者优先。

2. 有实时特征计算、特征平台(Feature Store)建设经验者优先。

3. 熟悉图数据库(如 Neo4j/HugeGraph)或图计算框架,有团伙作弊/黑产团伙挖掘数据处理经验者优先。

4. 有大规模指标一致性治理、端侧埋点治理及数据质量监控体系建设经验者

Job Responsibilities

1. Take charge of data infrastructure for core risk control businesses including account security, anti-cheating, marketing risk management and content security. Build the risk indicator system, and design & develop offline and real-time risk Feature Store.
2. Construct high-concurrency, low-latency real-time risk data pipelines to supply high-quality data for risk engines and algorithm models, including black sample library construction, device fingerprint data processing, and extraction of user behavior sequence features.
3. Formulate collection specifications for end-cloud security data (e.g., tracking logs, device environment, network features), conduct data quality monitoring and standardized governance, and guarantee the accuracy, completeness and timeliness of underlying risk data sources.
4. Realize precipitation and reuse of risk data assets, and ensure production data quality and SLA standards. Collaborate cross-functionally with risk strategy, algorithm and business teams to rapidly respond to data demands in the fight against underground fraudulent groups.

Job Requirements

1. Bachelor's degree or above in Computer Science or related majors, with 1+ years of big data development experience and proven track record of building real-time or offline data systems.
2. Solid programming foundation in Java / Scala / Python, proficient in SQL; capable of massive data development, complex logic processing and performance tuning.
3. Familiar with offline big data ecosystems such as Hadoop, Spark and Hive; expert in real-time computing frameworks like Flink, with practical experience in unified stream-batch architecture or high-throughput, low-latency real-time streaming projects.
4. Skilled in OLAP engines including ClickHouse and Doris; master data warehouse modeling theories such as dimensional modeling, with strong capabilities in data abstraction and architecture design.
5. Strong sense of responsibility and business acumen; able to deeply understand fraudulent tactics of underground industries and proactively identify business risks from a data perspective.

Preferred Qualifications

1. Experience in data warehouse development for business security, risk control or anti-cheating; familiarity with device fingerprint, account security systems or countermeasures against underground fraud rings is a plus.
2. Prior experience in real-time feature calculation and Feature Store platform construction is a plus.
3. Familiarity with graph databases (Neo4j, HugeGraph, etc.) or graph computing frameworks, with experience in data processing for detecting group cheating and fraudulent gangs is a plus.
4. Experience building large-scale indicator consistency governance systems, client-side tracking log governance frameworks and data quality monitoring platforms is a plus.