Programming Languages
PythonJavaScriptTypeScriptHaskellPrologPDDLR
Monica (Yiyang) Chen
Software Engineer & Data Engineer
Melbourne, Australia
I’m a Software & Data Engineer with a background in Bachelor of Data Science and Master of IT (Artificial Intelligence) from the University of Melbourne. I build and scale modern systems end to end, combining backend services, data pipelines, and production-ready practices.
I enjoy building backend features, integrating APIs, and supporting frontend workflows, contributing to production-ready systems through solid testing and CI/CD. On the data side, I work with machine learning models, NLP, CV, and LLM-based analytics, and I’ve built scalable ETL workflows and distributed pipelines that turn large datasets into useful insights. I also create clear data visualisations that help stakeholders make informed decisions.
Skills
Tech Stack
React.jsNext.jsNode.jsFlaskCSS3HTML5DockerGoogle Cloud Platform
Data Analytics
ETL developmentHadoopPySpark
Database
SQLMySQL
Cloud
AWSGoogle Cloud PlatformDockerCI/CD · GitHub Actions
Tools
TableauPower BIJiraFigmaGitHub Actions
Test
Unit testIntegration test
Languages
EnglishChinese (Mandarin)
Work Experience
Fujian Big Data Exchange Platform · Data Engineer
Fujian, China · Nov 2023 – Feb 2024
Description
Designed and deployed machine learning models to predict bid-winning amounts and award timelines, covering the full lifecycle from hypothesis through modelling, validation, and delivery.
Action
- Trained machine learning models using Python (Scikit-learn) and PySpark MLlib to forecast bid-winning amounts and award lead time.
- Manipulated Spark DataFrames to clean, transform, and engineer features across large, multi-source datasets, improving data quality for downstream modelling.
- Optimized model performance via feature engineering and tuning to provide quantitative, data-driven support for bidding strategy.
- Presented data insights and key facts to stakeholders, improving the decision-making process.
- Created visualised charts in Tableau dashboards and reports to demonstrate key features of large-scale data.
- Built automated preprocessing workflows on a Hadoop-basedenvironment, supporting scalable ETL, structured data handling, and cloud-aligned data engineering practices.
Achievement
Transformed data insights into forward-looking strategic decisions, shifting analysis from descriptive to predictive guidance.
WEHI · Full Stack Developer
Melbourne, Australia · Jul 2024 – Nov 2024
Repo
Description
Optimised the Student Organiser Platform by building data pipelines and automated testing workflows to improve system reliability and support data-driven development.
Action
- Built synthetic test datasets using SQL to expand test scenarios and significantly improve test coverage and system reliability.
- Developed and optimised Flask backend services to streamline data ingestion, processing, and API delivery, enhancing data flow efficiency across the platform.
- Containerised application components with Docker and implementedCI/CD pipelines to automate testing and deployment for consistent, reliable releases.
- Collaborated with cross-functional teams to diagnose system bottlenecks and deliver scalable, production-ready improvements.
GF Securities Co., Ltd. · Data Analyst
Nov 2024 – Feb 2025
Description
Supported data-driven decision-making by building reliable data pipelines, preparing ML-ready datasets, and delivering analytical insights through SQL, Python, and automated ETL workflows.
Action
- Extracted, transformed and validated business data using SQL and Python to support analytics and ML-ready dataset preparation.
- Built automated ETL workflows to improve data refresh efficiency and enable reliable, repeatable reporting processes.
- Performed EDA to identify key customer patterns and operational insights that inform business decisions.
- Developed clear data visualisations using Power BI to communicate trends and support stakeholder decision-making.
- Worked with modeling and risk teams to assemble, clean and quality-check datasets for predictive modeling use cases.
Projects
Projects
AI Recipe Assistant
GitHubEnd-to-end AI-powered recipe recommendation system combining retrieval, LLM tool-calling, multi-turn reasoning, and controlled session memory.
- Built a deterministic retrieval pipeline on SQLite to search and rank recipes from pantry ingredients, keeping recommendations grounded and explainable.
- Designed a hybrid architecture that combined rule-based filtering with LLM reasoning to improve candidate selection and reduce hallucinations.
- Orchestrated structured workflows and tool-calling agents with LangChain so the assistant could choose between direct answers, clarification, and backend actions.
- Implemented session-scoped conversational memory with explicit reset semantics to support personalization without uncontrolled context buildup.
- Integrated OpenAI models for intent parsing, multi-turn dialogue handling, and tool selection inside the agent framework.
- Developed backend scoring and normalization logic to measure cookability based on ingredient coverage and substitution feasibility.
ArtCert
GitHubArtwork authentication and provenance platform with secure data workflows and scalable backend architecture.
- Implemented Flask REST services with JWT security and Redis-backed caching to improve data access performance.
- Designed and optimised MySQL schemas to support high-integrity transactional workloads and efficient analytical queries.
- Built containerised services and automated CI/CD pipelines to deliver reproducible deployments on Google Cloud Platform.
- Partnered cross-functionally in Agile sprints to validate requirements, optimise data structures, and ensure scalable system performance.
CityBite
GitHubEnd-to-end analytics platform in R Shiny with R and Python workflows to analyse customer behaviour and optimise promotional spend.
- Collected customer and promotional data via a custom Google web-scraping pipeline; cleaned and feature-engineered the dataset in Python.
- Built customer segmentation models (clustering, classification) to identify high-value segments and estimate customer lifetime value for targeted marketing.
- Developed an R Shiny platform that automated behavioural analytics for continuous insight generation.
- Delivered data-driven recommendations to optimise promotional strategy and engagement, backed by statistical testing and predictive modelling.
- Shipped interactive R Shiny dashboards integrated with Tableau, maintaining reproducible modelling workflows with thorough documentation.
Climate Change Fact-Checking System
End-to-end ML pipelines for climate misinformation detection using deep learning, NLP, and computer vision.
- Processed 1M+ text samples and 5K+ visuals with automated NLP (cleaning, feature extraction) and CV preprocessing workflows.
- Engineered features in pandas DataFrames, tuning and normalising inputs to improve downstream model quality.
- Trained and fine-tuned Transformer and DNN models (TensorFlow, PyTorch) to boost classification precision.
- Used OpenCV and CNN layers to extract visual features and integrate them into multimodal classifiers.
- Generated LaTeX analytical reports with accuracy, F1, ROC, and Grad-CAM summaries to interpret model behaviour.
- Built reproducible ML pipelines with clear train/test splits and systematic validation aligned to production best practices.
Contact
Phone
+61 420 948 980