Zechu Liu

Machine Learning Engineer

Singapore, Singapore4 yrs 2 mos experience

Most Likely To Switch

Key Highlights

Expert in machine learning and NLP applications.
Proven track record in developing impactful algorithms.
Strong foundation in computer science and software engineering.

Stackforce AI infers this person is a Machine Learning Engineer specializing in E-commerce and B2C applications.

Contact

Skills

Core Skills

Machine LearningNatural Language Processing (nlp)Data Analysis

Other Skills

Cloud ComputingBig DataGolangWeb DevelopmentAndroid DevelopmentProgrammingiOS DevelopmentProject ManagementComputer VisionSwiftRuby on RailsFlutterPythonNode.jsJavaScript

About

Passionate machine learning engineer eager to learn and stay at the forefront of the ever-evolving machine learning field, with a particular interest in search & recommendation systems, powered with LLM. Work eligibility: Singapore PR

Experience

4 yrs 2 mos

Total Experience

1 yr 4 mos

Average Tenure

1 yr 11 mos

Current Experience

Tiktok

Machine Learning Engineer

Jun 2024 – Present · 1 yr 11 mos · Singapore · On-site

TikTok Search - Local Services
Built and launched two LLM-powered search summary products for local services results
Supervised fine-tuning (SFT) open-source LLMs to enhance output quality
Developed end-to-end pipelines for offline card generation and online retrieval/ranking systems
Optimized system performance to balance user experience, efficiency, and content quality
Developed and optimized CTR prediction models for video POI anchors under Local Services traffic
Improved ranking strategies for local services video content, improving result quality, timeliness, and location-based services (LBS)

Machine LearningNatural Language Processing (NLP)Cloud Computing

Shopee

Machine Learning Engineer

Jul 2022 – Jun 2024 · 1 yr 11 mos · Singapore, Singapore · On-site

Developed retrieval and coarse-ranking algorithms for ads recommendation system on Shopee’s
e-commerce platform. These solutions greatly increased business metrics (revenue, GMV, and order) across 9 countries on 2 major Discovery Ads placements.
Implemented two-tower deep vector retrieval model with contextual and personalised information, where the query tower extracts multiple interest vectors from user behaviour sequence.
Enhanced coarse-ranking pCTR and pCR DNN model (based on ESMM) by training the model with more valuable features.
Designed and optimised several heuristic retrieval queues using collaborative filtering and other rules-based algorithms.
Processed big data with Spark on HDFS cluster for data analysis, training sample generation, feature generation and producing offline retrieval queues.
Improved and maintained supporting infrastructure and pipelines in Golang, such as retrieval and coarse-ranking visualisation tools, greatly accelerating our team’s development process.

Machine LearningData AnalysisBig DataGolang

Cloudflare

Solutions Engineer Intern

May 2021 – Aug 2021 · 3 mos · Singapore, Singapore

Helped the team better serve customers through building proof-of-concept solutions.
Implemented DDoS simulation tool.
Set up virtual private clouds (VPCs) on different cloud platforms and configured Generic Routing Encapsulation (GRE) tunnels to connect them via Cloudflare.
Built web applications on serverless platform.
Conducted webinars on technical deep dives into products.

Cloud ComputingWeb Development

Hope technik

Software Engineer Intern

Jan 2021 – Apr 2021 · 3 mos · Singapore

Built a REST backend server that integrates with a robot that runs on ROS.
Developed an Android tablet application that interfaces with the robot, with functionalities including WebRTC video streaming, joystick manual control, robot status update, biometrics scanning with a fingerprint scanner.

Web DevelopmentAndroid Development

Visa

Software Engineer Intern

May 2020 – Jul 2020 · 2 mos · Singapore

Interned under Digital and Developer Platform team.
Developed solutions that leverage Visa payment APIs to mitigate the impact of Covid-19 for cardholders, and implemented an Android application with Bluetooth Low Energy (BLE) that provides diners with a seamless and contactless experience of ordering and paying.
Worked with teammates in Intern Case Challenge to develop a solution to accelerate cash displacement.

Android DevelopmentWeb Development

National university of singapore

Undergraduate Teaching Assistant

Aug 2019 – Dec 2019 · 4 mos · Singapore

For CS1101S (Programming Methodology I)
Conducted tutorials and consultation sessions.
Prepared teaching materials.
Graded assignments.
Received 4.8/5 mean rating in anonymous feedback and 4 students’ nominations for teaching awards.