Yufei G.
End-to-End Custom LLM Application
Experience
Jul 2025 - Oct 2025
4 monthsEnd-to-End Custom LLM Application
G.A.I.A. (Game AI Assistant)
- End-to-end local LLM engineering: Engineered a vertical domain LLM application locally, orchestrating the full lifecycle from synthetic data generation and QLoRA fine-tuning (Qwen 1.8B) to GGUF quantization and deployment via Ollama.
- Advanced RAG & prompt engineering: Designed a robust retrieval system using LangChain and ChromaDB. Applied advanced prompt engineering strategies (e.g., Chain-of-Thought, Few-Shot Prompting) to align the small model's reasoning capabilities and significantly reduce hallucinations in domain-specific tasks.
- Full-stack app development: Developed a production-ready frontend using Chainlit, featuring real-time token streaming and visualization of the reasoning process (CoT) to enhance user interaction and system transparency.
Dec 2024 - Jun 2025
7 monthsSingapore
Quality Assurance Engineer (Web, iOS, Android, Java Backend)
Binance
- Conducted deep-dive debugging by tracing distributed logs and verifying data consistency across MySQL and Redis, ensuring that complex transaction states (e.g., Pending vs. Locked) were accurately synchronized.
- Expanded the Java-based automation suite (TestNG) by engineering robust test scripts for 20+ new payment microservices, increasing regression coverage by 30% for critical financial flows.
- Managed daily release trains via Jenkins, orchestrating canary deployments that routed live traffic to 100,000+ users per rollout, ensuring zero downtime during high-traffic windows.
- Validated end-to-end crypto-fiat settlement logic, rigorously testing edge cases such as network timeouts, double-spending prevention, and idempotency to guarantee 100% financial accuracy.
Mar 2024 - Dec 2024
10 monthsSingapore
Full-Stack Developer (Python, AWS)
Starstruck Pte Ltd
- Engineered high-concurrency web crawlers (Python/Selenium) capable of traversing 300+ global marketplaces, successfully processing over 120,000 data points daily with 99.9% uptime.
- Directed the creation of data workflows for 120,000+ records across 2,000+ companies, deploying automated validation and testing pipelines on AWS Lambda and EC2, achieving 40% higher accuracy in data handling.
- Established cloud infrastructure workflows for seamless data orchestration from S3 to DynamoDB, and implemented Spark ETL pipelines to optimize Big Data workflows, which lowered system latency by 20%.
Aug 2022 - May 2023
10 monthsUnited States
Backend Developer (Java)
Trending Lines – USC Daily Trojan
- Deployed scalable backend architecture with Spring Cloud and Docker, enabling multi-user access and consistent multi-node deployments.
- Optimized data retrieval by 30% and responsiveness by 50% through Elasticsearch tuning, Redis caching, and distributed query enhancements.
- Implemented real-time processing with KafkaStream for timely notifications and data computation, improving overall user experience.
Summary
A results-oriented software engineer with a Master's in Computer Science and hands-on experience in the end-to-end development of LLM applications. Proven ability to build, fine-tune, and evaluate custom AI solutions on cloud platforms like Vertex AI. Seeking to leverage strong backend engineering skills to create scalable and reliable AI-powered products.
Languages
Chinese
NativeEnglish
AdvancedEducation
Jan 2024 - Jun 2025
Singapore Management University
MSc in Management · Management · Singapore · 4.00/4.0
Aug 2021 - Jul 2023
University of Southern California
M.S in Computer Science · Computer Science · Los Angeles, United States · 3.77/4.0
Sep 2019 - Jul 2021
University of California Irvine
B.S in Computer Science · Computer Science · Irvine, United States · 3.89/4.0
Need a freelancer? Find your match in seconds.
Try FRATCH GPT More actions
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.