About me
Hi, my name is Florence. I am a Software Engineer based in the SF Bay area. My career interests lie at the intersection of full-stack development and applied machine learning/AI systems. Having built enterprise-scale web applications (React/TypeScript, Spring Boot) and data-intensive services (Druid, Redis) at Salesforce, I’m now actively expanding into ML/AI engineering. I aim to leverage my expertise in distributed systems and real-time data processing to ML infrastructure for production-grade AI applications.
Experience
- Senior Full-Stack Software Engineer - Salesforce Inc. (Aug 2021 - Present)
- Led the front-end development and launch of the Metering Usage Interface, utilizing React, Redux, Redux-Saga, and TypeScript. Successfully integrated a new API with Druid, enhancing data ingestion and querying capabilities. This pivotal upgrade significantly improved product development and backend efficiency, directly supporting enterprise-level deals and facilitating the closure of contracts worth millions. Ensure the new UI is tailored to meet the sophisticated needs of enterprise customers, underlining Salesforce’s commitment to providing scalable and high-impact solutions.
- Designed and implemented Metrics and Distributed Tracing UI with React and TypeScript, with HighCharts library integration delivering an intuitive user experience for system diagnostics and leading to improved usability and customer satisfaction. Enable real-time updates and high scalability for monitoring services, accommodating over a million services
- Developed and expanded the Metering, Metrics, and Traces services API using Spring Boot and Redis, which broadened customer access to essential app services and contributed to an increase in user engagement.
- Software Engineer - Fivetran Inc. (Jul 2019 - Sept 2020)
- Developed, maintained the data pipeline using Java for File services with 10+ billion monthly syncing rows
- Designed and implemented a new file pattern-detection algorithm for S3 listing phase to speed up the source extraction time by 20% for enterprise customers, integrated with AWS Java SDK
- Designed and built the REST API login task flow with Java, Google Cloud Datastore, and Okta OpenID protocol for the company’s internal debugging tool to improve data privacy practices, gain trust in enterprise business
- Full Stack Software Engineer - Yewno Inc. (July 2018 - Apr 2019)
- Developed the Front-End and User Interface of Yewno Edge app with JavaScript, ES6, React, React-Router, Redux, Redux-Saga, and Immutable.js
- Co-authored the subscription system with Stripe React.js for front-end and Python for back-end API
- Utilized Ant Design React UI library for animation and special effects to increase users experience
- Worked closely with the team to optimize existing API calls in Python; refactored and deprecated unneeded API calls
- Applied React High-Order Component to improve component reusability
- Implemented Snapshot testing with Jest along with Enzyme
- Yewno Edge demo: https://www.linkedin.com/company/yewno/
AI Certifications
AI Projects
House Price Prediction | Kaggle Competition | 2025 on going
- Optimized data pipeline: Handled missing values (median imputation), encoded categorical variables (target encoding)
- Tools: Python, Pandas, Scikit-learn, XGBoost
Feature project
Layr is a decentralized, peer to peer file storage system built atop Kademlia DHT that enforces data integrity and availability, with cryptocurrency-based incentive scheme.
Quick demo for uploading and downloading a short movie:
Quick demo for auditing and patching a redundant shard of copy:
Personal projects
- Construction site: React project on Vercel:
Live here - MyFlix: A Netflix-like video subscription application built with Rails
Live here GitHub Link
- PetHome: A social network site dedicated for pet lovers built with Rails
Live here GitHub Link
- FloTrello: a project management tool allows you to drag & drop the cards and lists in the board
Live here GitHub Link
- Restaurant Site: a front end project for ordering sushi
Live here