Daisy · Portfolio

Hi, I’m Daisy (Yitong) Pei 👋

AI Enthusiastic | Software Engineer | Automation Test Analyst
I build AI-powered applications and GenAI features, integrating automation testing to deliver reliable, production-ready solutions.
React
JavaScript
Playwright
Node.js
RAG/LLM
Python
LangChain
Daisy Pei
Preview

About

I am a result-driven Full Stack & AI Engineer with a Master’s in AI and a Bachelor’s in Computer Science, experienced in delivering AI-powered applications, automation testing frameworks, and scalable web/mobile solutions. Skilled in LLMs, and RAG to digitise workflows, reduce manual effort, and enhance user experience, with strong skills in Agile/DevOps, stakeholder communication, and cross-functional collaboration.
LocationMelbourne, Australia
FocusSoftware Development, Playwright Automation Testing, LLM, RAG, LangChain
EducationBSc Computer Science at the University of Melbourne, M.Sc AI at the Nanyang Technological University Singapore

Highlights

  • Shipped an on-prem Local RAG stack (LangGraph, FAISS, BM25 fusion, bge-small-en embeddings, citation-only prompts) with regex redaction and keyword red-flag guardrails; reduced document lookup time by 70%.
  • Engineered a policy-aware Donor Eligibility Checker with policy-tagged chunks, role-based retrieval filters, and audit logs (who-saw-what-and-why);
  • Built and maintained Playwright regression/E2E suites (fixtures, network mocking) wired into CI/CD with parallel sharding and flaky-test quarantine, boosting release reliability.
  • Delivered full-stack features across React, React Native, Node.js, and Flask/FastAPI; typed APIs via OpenAPI, JWT-based auth, and streaming responses (SSE/WebSocket) for LLM flows.
  • Implemented ML workflows: data prep in pandas, prompt/RAG evaluation with golden Q&A sets, packaging inference endpoints (FastAPI/uvicorn) and trials on AWS SageMaker.
  • Built a Power Apps estimation tool (anonymous planning-poker) and manage the Power Platform CoE (governance, DLP policies, ALM) used across teams.
  • Optimized donor app performance in React Native (batched API calls, memoization, list virtualization), cutting screen load times by 25%.
  • Comfortable with MongoDB/MySQL, REST/WebSocket integrations, Git workflows, and Agile/DevOps practices; mentor graduates and run internal workshops.

Certificates and Awards

  • AWS Certified Machine Learning Associate (2025)
  • AWS Certified AI Practitioner (2025)
  • Unimelb - Melbourne Plus People Leadership (2023)
  • Unimelb - Global Mobility Scholarship (2022)
  • Unimelb - Melbourne International Scholarship (2021)

Projects

Total Projects

9

AI Projects

5

© 2025 Daisy (Yitong) Pei. All rights reserved.