Senior Data Engineer - AI Data Service at Binance

This is a specialized engineering role supporting Binance Square, the platform's community and content feed (similar to a "Twitter for Crypto" embedded in the exchange).

Unlike a general Data Engineer who might focus on BI dashboards, this role is deeply integrated with Machine Learning. You will be building the infrastructure that feeds the Recommendation and Search algorithms. Your primary customer is the algorithm team; your product is "Features" (clean, processed data variables) delivered in both real-time (online) and batch (offline).

Key Responsibilities
Feature Engineering Infrastructure: Design pipelines that transform raw user behavior data into features for ML models (e.g., "User clicked X times in the last hour").

Pipeline Architecture: Manage high-scale ETL/ELT pipelines using distributed systems.

Online/Offline Consistency: Ensure that the data used to train models (offline) matches the data available when the model makes a prediction (online/real-time).

Optimization: Monitor data quality and stability to prevent "garbage in, garbage out" scenarios for the AI models.

Technical Requirements
Experience: 5+ years in Data Engineering.

Languages: Java and Scala are mandatory (Expert level). This suggests a heavy reliance on JVM-based Big Data tools.

Stack: Spark, Hive, Flink (likely for real-time streaming).

Concepts: Feature Stores, Data Warehousing, Machine Learning Lifecycle.

Language Skills: Must be Bilingual in English and Mandarin to coordinate with global and HQ-based teams.

Culture & Environment
Remote-First: Binance is famous for its decentralized structure. You are expected to work autonomously without a physical office.

Pace: "Hardcore" is a term often associated with Binance's internal culture. It is fast-paced, high-pressure, and results-oriented.

Compensation: While not listed, Binance is known to pay above-market rates, often with the option to receive portions of salary in cryptocurrency (USDT/BNB).

Diversity in Crypto Engineering
Binance does not publicly release granular diversity reports. However, the intersection of Blockchain, Fintech, and Data Engineering generally reflects the following industry benchmarks:

Gender: Heavily male-dominated (Global estimates for crypto engineering roles often sit below 10% female representation).

Geography: Highly distributed. The "Asia/Pacific" focus of this role aligns with a significant portion of Binance's engineering talent pool.

Strategic Insight: The "Feature Store"
The job description mentions "offline training features and online serving features." This is the critical technical challenge of the role.

Offline: Using historical data (via Spark/Hive) to train a model.

Online: Using real-time data (via Flink) to serve recommendations instantly when a user opens the app.

The Challenge: Making sure User_Click_Count is calculated exactly the same way in both environments to avoid "Training-Serving Skew."