Sicong E

Global architecture and design firm integrating data-driven methodologies for enhanced decision-making

• Designed and implemented an end-to-end data pipeline orchestrated by Apache Airflow to manage over 100,000 evolving Building Information Modeling (BIM) data points for 6,600 curtain wall units of Tencent’s 207,700 m² Global Headquarters, achieving a 75% reduction in manual effort and a 50% efficiency gain
• Automated ETL workflows integrating Autodesk Platform Services APIs for real-time metadata extraction, Pandas for data pre-processing, and AWS SDK for Python to load raw data into Amazon S3 and trigger AWS Glue jobs
• Engineered advanced transformations in AWS Glue using PySpark scripts, including timestamp-based partitioning and schema cataloging, enabling seamless integration with Amazon Redshift Spectrum for downstream analytics
• Delivered actionable insights by leveraging Redshift Spectrum for querying and visualization, supporting construction detail modifications, cost optimization, and compliance with energy and fire safety standards

Architecture firm specializing in retail, mixed-use, residential, and large-scale urban planning

• Led architectural design and BIM documentation for 10+ large-scale urban projects, collaborating with clients and consultants. Explored Python automation to streamline workflows, sparking a transition into data engineering

This program equipped me with the skills to advance my career in data engineering. Key learning outcomes include:

• Data Processing, Analysis and Visualization: Proficient in Python and multiple libraries, including NumPy, Pandas, Matplotlib, PySpark, SQLAlchemy, PyMongo, Scikit-learn, SciPy, DASK, NLTK, Graphviz, and Paho.
• Database Design and Management: Proficient in SQL; wrote complex database queries; conducted database schema design (data modeling); involved common RDBMS and NoSQL databases; applied regular expressions.
• Database Containerization: Used Docker to create and manipulate images and containers; performed change data capture (CDC) in different types of databases including MongoDB, Cassandra, Redis, and Firebase.
• ETL Processes and Data Pipelines: Performed extract, transform, and load (ETL) operations using Python; created ETL pipelines and orchestrated workflows using Apache Airflow.
• Big Data Handling: Processed big data using Spark and Hadoop; ran parallel operations using DASK; handled real-time streaming data using Mosquitto, ThingsBoard, and Kafka.
• Statistics Basics: Identified key concepts in statistics, including various types of probability distributions, Central Limit Theorem, and correlation.
• AI/ML Algorithms: Built a prediction modeling using linear regression; implemented foundational ML algorithms, including gradient descent for error reduction, classification with Naïve Bayes theorems, clustering (k-means), reinforcement learning (Q-matrix, Bellman equation), and deep neural networks.
• Software Engineering and Network Basics: Grasped commonly used command line commands; identified key concepts of Java, asynchronous event-driven programming, and HTTP and client–server architecture; created applications using Flask web server and Jinja templating language.

USC provided a comprehensive education in architecture, emphasizing data-driven aspects. This honed my ability to manage and analyze complex datasets in building design and construction. Key learning outcomes relevant to data engineering include:

• BIM Data Management: Proficient in Building Information Modeling (BIM) using Revit, focusing on managing and analyzing large datasets generated by BIM processes.
• Data Analysis in Architecture: Skilled in extracting and manipulating BIM data to enhance project decision-making and optimize design and construction processes.
• Revit Proficiency: Expertise in using Revit for architectural design and data extraction, contributing to efficient project workflows and accurate data management.

This unique combination of skills positions me well for roles in data engineering, where technical proficiency and the ability to manage and interpret large datasets are crucial.