Carl Bai

As a passionate Software Engineer and an admirer of tech, I would like to apply for the Software Engineer position. I feel confident and have a solid software development background as a potential Software Engineer.

Meta Platforms, Inc., Bellevue, WA, USA

         Production Engineer Intern                                                                                                                   May 2022 – August 2022

• Worked on a NewSQL database that is leaderless, ACID compliant, and auto-scalable with a team of 20+ Software Engineers and Production Engineers.
• Built a web-based instrumental service to facilitate on-call operations for a team of 200+ people. Used C++/Python for backend and React.js, Tailwind CSS, HTML & JavaScript for frontend. Implemented features like search support for AJAX query, pagination support for cluster inspection, and quick access.
• Delivered multiple operational tooling improvements. For example, improved CLI startup time from 13 seconds to 0.25 seconds (98 % improvement) using Lazy Import Strategy; utilized Pagination Support for APIs on the backend, and reduced result inspection time from 6.83 seconds to 0.863 seconds (87.3 % improvement); fixed the task crush bugs for the whole Storage Team on the on-call operations by modifying the Hybrid Check Mechanism.
• Prototyped the visualization charts for failure domain, capacity usage, and latency graph for network connection status of nodes to present more statistics and user-friendly information about the Clusters, highly enhanced user experience and work efficiency by 83.6%.
• Overhauled the unit tests and significantly increased testing coverage for the instrumental service to support TDD (Test-Driven Development).

Beijing Thunisoft Corp. Ltd., Shanghai, China (One of the top 50 Software Companies in China)

         Software Engineer                                                                                                                   December 2020 – July 2021

• Led a team to design and develop a client platform, worked on frontend user interface and backend cloud service using Python, Java, TensorFlow, FastAPI, AngularJS, and REST APIs, allowing users to select from 20 speakers to generate speeches with Text, Speed, and Pitch.
• Wrote Shell scripts and Linux commands to build, maintain, and deploy both development and production environment, using Docker Images/containers, Mesos and Kubernetes, and load-balancing tool Nginx to validate portable deployment among multi-servers and concurrent external connections.
• Initiated Mockito with Java Spring Framework based on AWS for mocking framework, fixing bugs, tracking analysis and alleviating the workload by above 80%.
• Originated the Service Oriented Architecture (SOA) application using JMS, Spring IOC, Spring MVC Framework, and Spring Batch, and handled the security using Spring Security to develop fully automated client TTS systems, significantly improving maintenance efficiency of specific users' client accounts by 78.6%.
• Enforced an innovative approach to oversee the transition of the commerce APIs to microservices that running on a dockerized platform and ECS through other Amazon Web Service (AWS) such as Dynamo, SQS, and ElastiCache Redis that successfully reduced the operation costs by over 75% upon the newly service.
• Modularized end-to-end TTS speech synthesis system based on Conformer-based FastSpeech2 and X-Vector, realized the speech synthesis for multi-speakers and delivered features such as voice tweak with speed and pitch. Obtained MOS up to 4.5 and improved the previous state by 1 MOS.
• Trained and reconstituted Integration with neural vocoders from Parallel WaveGAN to Multi-band MelGAN and significantly reduced RTF of synthesis from 2.76 to 0.18, based on 300 bytes text and CPU of Intel(R) Xeon(R) CPU @ 2.20GHz.
• Optimized backend Feature Processing by utilizing the ML-based approach and enhanced the word and tone pronunciation accuracy rate to above 99%.

Huimei Healthcare, Shanghai, China (A healthcare group co-founded by Mayo Clinic and Hillhouse Capital)

         Software Engineer & Data Scientist                                                                                                   January 2020 – December 2020

• Led a team of engineers, PM, & UX for the BMS Social Listening Project, responsible for tracking 23.6 million medical data from 28 different medical platforms.
• Developed on Spark using Scala API and SparkSQL for faster interactive queries, streaming data processing of data, and Integration with MongoDB with a massive volume of millions of data, using Spark API over Hortonworks Hadoop YARN in Hive, which vastly improved data processing performance by 86.7%.
• Extracted data from web crawler and extensively used aggregation operation APIs in MongoDB and regular expression & Pandas in Python for sentiment analytics to filter, merge, and transform related data, then loaded converted data from the staging area to ElasticSearch.
• Utilized Kibana to visualize the medical data from ElasticSearch and used python-based Lucene query syntax, PyLucene, to filter the specific data, implemented ETL Filtrating step using Logstash-Kafka-Logstash among various microservice modules, boosting the search efficiency by over 90% compared to MySQL.
• Deployed the sentiment analysis model trained on HM-BERT (advanced BERT method based on original Google BERT method) to analyze the emotional Text for patients' medication tendency and achieved 91.8% F1 scores, improving the Google BERT method by 2.3%.
• Configured and built Alibaba Cloud Web Hosting Service, including dependency, driver, account, domain name, and permission allocation.

Huawei Tech CO., LTD., Wuhan, Hubei, China        

Software Engineer Intern                                                                                             January 2018 – July 2018

• Worked on the team that builds FusionCompute, focusing on tuned high-performance and high reliability in VM instance provisioning, clustered resource pool management, and intelligent HA/FT scheduling.
• Conducted E2E operation of Huawei cloud computing and performed E2E O&M in HCIE-Cloud.
• Designed, built, and maintained the infrastructure of OpenStack and software platforms within DevCloud using infrastructure-as-a-service and platform-as-a-service offerings for public cloud service. 
• Managed OpenStack infrastructure upgrades, expansion, scaling, troubleshooting, and debugging for 20 clients per week.

EDUCATION  

University of Southern California (USC), CA, USA                                                                                                                            August 2021 – May 2023(Exp)

Master of Science in Computer Science

• Major: Computer Science          GPA: 3.85 
• Web-based instrumental Yelp-Seek service

University of California, Riverside (UCR), CA, USA                                                                                                                     August 2018 – June 2019

Bachelor of Science in Computer Science

• Major: Computer Science                GPA: 3.89
• MIT Remote AI Chatbot Research

Wuhan University of Technology (WHUT), Hubei, China                                                                                                            September 2015 – June 2019

Bachelor of Science in Software Engineering

• Major: Software Engineering                 GPA: 3.9
• Honors & Rewards: First-Class Scholarship (2017); "Excellent Student" Reward (2017); First-Class Scholarship (2018)