01 · INPUT
QUERY RECEIVED
AI & Junior Engineers · 2025–2030
02 · AGENT
PLANNER AGENT
gpt-4o-mini · 5 queries planned
04 · SYNTHESIS
DRAFT WRITER
1000+ words · 9 sections
05a · EVALUATOR
CLAUDE SONNET
score: —
GAP ≥ 4 · CONSERVATIVE MIN
05b · EVALUATOR
GEMINI FLASH
score: —
06 · TARGETED
EXTRA SEARCH
fill weak sections
07 · REWRITE
SECTION REWRITER
targeted rewrites
08 · GUARD
BEST-VERSION GUARD
rollback if score regresses
09 · OUTPUT
CONVERGENCE
quality score 0/10
INITIALIZING PIPELINE...
Deep Research Agent Pipeline

Shuangyi
Hu

Building AI systems that think, search, evaluate, and refine — end to end.

View Projects Get in touch
Scroll
✦Java ✦Spring Boot ✦Distributed Systems ✦React ✦Node.js ✦Redis ✦RabbitMQ ✦Docker · AWS ✦Gemini AI ✦LLM Pipelines ✦Microservices ✦Full-Stack Dev ✦Java ✦Spring Boot ✦Distributed Systems ✦React ✦Node.js ✦Redis ✦RabbitMQ ✦Docker · AWS ✦Gemini AI ✦LLM Pipelines ✦Microservices ✦Full-Stack Dev
Shuangyi Hu
Shuangyi Hu
M.S. Computer Software Engineering · Northeastern
● Open to work GPA 4.0 Seattle, WA

About me

I build systems that scale without breaking and interfaces people actually want to use. Passionate about the intersection of backend rigor, distributed architecture, and AI-augmented workflows.

🎓
Education
Northeastern University · GPA 4.0
💼
Status
Open to work · SWE / Full-Stack / AI
📍
Location
Seattle, WA · Open to Remote & Relocation
🌐
Languages
Mandarin (Native) · English (Professional)
⚡
Core Stack
Agentic Workflows · Python · Java · JavaScript · React · AWS
📧 Email Me 💼 LinkedIn 📄 Resume
01

Selected Projects

↗
💚 Why spend hours searching when a team of agents can read, evaluate, and disagree for you?

AI · Multi-Agent Pipeline

AI Research Agent

Multi-agent pipeline with parallel retrieval, dual-model evaluation (Claude + Gemini), and a best-version guard — iteratively refines until quality threshold is reached.

Python Claude API Gemini API OpenAI SDK
↗
💚 For people who believe the best places are meant to be shared.

Full-Stack · MERN

WanderMark

Place bookmarking platform with JWT auth & Cloudinary uploads. Reduced API response time by 87% and P95 latency by 30% via async uploads and indexed queries.

React Node.js MongoDB Cloudinary
↗
💚 Your inbox shouldn't need a second job to manage.

AI · Chrome Extension · Hackathon

MailMind

Led team of 4 to build an AI Chrome extension for intelligent email classification & summarization. −60% redundant API calls, UI responsiveness 1.2s → 400ms.

React TypeScript Gemini AI Chrome APIs
↗
💚 Because a missed booking or double-charge shouldn't be anyone's problem.

Distributed Systems · Backend

DrivePro

Microservices chauffeur platform with real-time booking, Redis caching, and Seata distributed transaction consistency. 1,000+ req/sec, Dockerized on AWS EC2.

Spring Boot Redis RabbitMQ AWS EC2
02

Skills &
Background

"I'm drawn to problems where scale demands creativity — where the architecture is as important as the code."
Java Python JavaScript TypeScript Spring Boot React Node.js Redis RabbitMQ Seata Docker AWS MongoDB MySQL Gemini AI Microservices LangChain
Education
Northeastern University
M.S. Computer Software Engineering
Seattle, WA · 2023–2025 · GPA 4.0
Currently Exploring
LLM integration & AI-augmented pipelines
Cloud-native distributed architecture
Developer experience & tooling
Product-minded engineering
Languages
Mandarin Chinese (Native)
English (Professional)
03

Let's Build
Something.

Open to full-time SWE roles and internships in backend, full-stack, or distributed systems. Always happy to chat.

📧 amandashuangyihu@gmail.com 💼 linkedin.com/in/shuangyi-hu
Currently available
Ready for new
challenges.

Based in Seattle · Open to remote & hybrid roles · SWE / Backend / Full-Stack · Willing to relocate

💬 Career Q&A · Powered by AI
Shuangyi
My Career Assistant
Career Conversation
🤖
Hi! I'm Shuangyi's career assistant 👋 Ask me anything about her skills, projects, or availability!