Welcome

Intro

Hello there! I'm Chun Lu, a software engineer driven by a passion for solving problems. Since graduating with a Bachelors in Computer Engineering from Stony Brook University in May 2018, I've gained experience in both automation and full-stack development. I'm particularly interested in learning about new technologies and how they can be incorporated into my workflow. I'm always looking for new challenges that allow me to leverage my skills and make a positive impact.

Work

Throughout my software engineering journey, I've constantly sought opportunities to expand my skillset and tackle new challenges. Here, I'd like to highlight some projects that allowed me to deepen my skills and knowledge, allowing me to develop the skills to solve problems creatively and efficiently.

SD-WAN Network Automation

SD-WAN (Software-Defined WAN) offers a dynamic network solution. It intelligently routes traffic across diverse connections (MPLS, internet, etc.) using software, not fixed configurations. This translates to increased flexibility, improved application performance by choosing the fastest path, and potential cost savings through affordable options. Plus, SD-WAN simplifies management with centralized control. It's a smarter way to connect your geographically spread locations.

While SD-WAN offers significant benefits like improved network performance and flexibility, manually configuring each device can be a significant bottleneck. Imagine a network administrator tasked with individually configuring hundreds of devices across geographically dispersed locations. This repetitive, time-consuming process is not only inefficient but also increases the risk of human error. A single mistake in a complex configuration can lead to connectivity issues, service disruptions, and even security vulnerabilities across the entire network.

Thankfully, automation tools can streamline the SD-WAN configuration process. By leveraging my expertise in Ansible, Python, YAML, Jinja, and vendor-specific APIs, I developed an automation solution that drastically reduced configuration times.

One hurdle in automating SD-WAN configuration was the lack of a unified data model representing the network. This model would provide the automation pipeline with the necessary configuration values for each device.

To address this challenge, Nautobot was used as the network source of truth. Nautobot, built using Django, includes a powerful application called Design Builder, which allows different networks to be modelled. I partnered with the network administration team to design the network using Nautobot's Design Builder application. This collaborative effort resulted in a centralized database of configuration data, expediting configuration and ensuring consistency across the network. This data could then be easily queried using GraphQL by the automation scripts. By leveraging automation tools and a centralized data model, I significantly reduced configuration time and ensured error-free deployments, ultimately improving the overall efficiency and reliability of the SD-WAN network.

Unmanned Surface Vehicle

Medium Displacement Unmanned Surface Vehicles (MDUSVs) navigate the seas without a crew. These high-tech vessels rely on GPS, radar, and sensors to chart their course, making them ideal for long-duration missions. From collecting ocean data to assisting in search and rescue, MDUSVs offer a versatile and autonomous approach to tasks at sea.

Extensive at-sea testing of MDUSVs is crucial for software refinement and bug elimination. These tests generate vast amounts of data logs (hundreds of gigabytes per event) from multiple sensor systems, critical for understanding the automation's decision-making process. However, not all data is valid or complete, and existing tools hindered our data scientists' ability to efficiently remove corrupt data and aggregate the valuable insights for analysis.

To address the challenge of processing massive datasets generated during MDUSVs testing, I co-developed an internal suite of data analysis software using Python and libraries like Pandas, NumPy, and Matplotlib. This software tackles the inefficiency of existing tools by ingesting data logs, intelligently removing corrupted entries, and aggregating the cleaned data for streamlined analysis. It also generates graphs and time series visualizations, empowering data scientists to quickly identify periods of interest for further investigation. This internal tool streamlines the data analysis process, allowing data scientists to focus on extracting valuable insights from the wealth of information collected during MDUSVs testing. This experience highlights my ability to develop data analysis solutions using Python and my understanding of the challenges faced in big data processing.

Digital Thread

Digital Thread Additive Manufacturing (DTAM) streamlines additive manufacturing (3D printing) by creating a single, connected flow of data throughout the entire process. This "digital thread" links design, modeling, printing, and even post-production, ensuring everyone involved has access to the latest information. This leads to improved efficiency, better quality control, and a smoother workflow for complex 3D-printed parts. DTAM allows manufacturers to trace materials, track progress, and optimize production for a more robust and data-driven approach to additive manufacturing.

DTAM's system lacked a secure platform for collaboration and file sharing, posing a challenge for authorized users working with 3D components. Additionally, they encountered difficulties in packaging specifications and 3D files into a single, unified PDF for easy reference.

I helped design and develop a full-stack 3D file sharing website using React, DRF, and Postgres. This platform allows authorized users to share 3D files, along with corresponding part specifications, facilitating collaboration and streamlined file management. Furthermore, I championed the development of an innovative feature: interactive 3D PDFs for component visualization. These PDFs leverage custom LaTeX templates and embedded U3D models, taking advantage of LaTeX's capabilities to generate dynamic 3D content within the document, ultimately enhancing user experience and accessibility. This experience honed my full-stack development skills and solidified my passion for exploring new technologies to enhance user experience.

Global Positioning and Navigation

Global Positioning and Navigation Systems (GPNS) rely on the NAVSTAR GPS, a space-based system that continuously delivers global position, velocity, and time data. NIWC Pacific supports the Navigation Systems Program by providing engineering expertise and conducting research to ensure the system's functionality and explore future capabilities.

While meticulous testing and data analysis are paramount for accurate positioning in satellite navigation systems, GPNS software often relies on third-party tools for data validation, creating a time-consuming and expensive process.

I joined a satellite testing software team for a three-month project to address this inefficiency by implementing real-time data analysis. Recognizing the potential for streamlining the workflow, I quickly mastered the system's C# and XML architecture. Leveraging navigation message documentation, I successfully integrated real-time analysis, eliminating the need for post-processing test log data. This not only reduced the validation time by [mention specific % or time saved] but also improved data precision by eliminating rounding errors introduced during post-processing. This project provided a valuable opportunity to rapidly acquire new skills in C# and XML, while delivering a lasting improvement to the software's capabilities.

LinkNYC

LinkNYC transforms New York City's streetscapes by replacing payphones with tech-driven kiosks. These all-in-one hubs provide free, super-fast Wi-Fi, keeping residents and visitors connected. The kiosks also offer free US phone calls, device charging stations, and a touchscreen interface for accessing city services and information. LinkNYC aims to bridge the digital divide and create a more connected, informed Big Apple.

Paper-based inspections created a time-consuming hurdle for LinkNYC kiosk installations, hindering both pre-installation suitability assessments and post-installation functionality verification.

To address this inefficiency, I co-developed an Android mobile application using Java, MongoDB, and Mongoose. This app streamlined the inspection process by digitizing forms and facilitating faster data capture, leading to a significant reduction in deployment times. My data management expertise further enhanced the solution, as I queried LinkNYC data via REST APIs to populate the app and transformed information into user-friendly CSV formats for easier management and updates. Furthermore, I created clear product lifecycle diagrams using Microsoft Visio, ensuring efficient project planning and execution. This project not only demonstrates my technical skills in mobile app development and data management, but also my ability to identify problems and contribute to efficient project planning.

Skills

Programming Languages

• Python
• Java
• Javascript
• C#

Front-End

• React
• Bootstrap
• CSS
• HTML

Back-End

• Django
• Django Rest Framework
• Flask

Databases

• PostgreSQL
• MongoDb
• Redis

DevOps

• Ansible
• Docker
• Podman
• GitLab CI
• GitHub Actions

Additional Tools

• Git
• AWS
• REST
• GraphQL
• Poetry
• YAML
• Jinja
• Node.js
• SQL
• Jira

Contact

Thanks for taking the time to learn more about me! I'm always interested in connecting with other passionate individuals in the tech industry. My contact information and links are listed below. I'm eager to hear about your projects and explore potential collaborations.

• LinkedIn
• Resume
• GitHub

Elements

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';