About Me

AI Engineer passionate about building robust, human-aligned AI through the interplay of knowledge representation, retrieval-augmented generation, and agentic systems.

Background

I'm Zhiruo (Zoey) Zhang, a Master's graduate in Informatics from the University of Zurich, specializing in Data Science. I began my journey with a Bachelor's in Software Engineering, exploring NLP research through my thesis, and later deepened my expertise with advanced courses and projects during my Master's studies.

During my Master's, I conducted research in Knowledge Graph Question Answering (KGQA) as part of my thesis and then worked as a Research Assistant. I kicked off my industry experience at an AI startup, where I quickly proposed and prototyped the GraphRAG project, demonstrating initiative, research-to-application skills, and the ability to address real-world problems in industrial settings.

I also worked on an early-stage startup project with friends, where I handled the entire technical stack while participating in product positioning, fundraising preparation (business plan, investor outreach), market research, and beta testing guidance. Though we didn't formalize the company at pre-seed stage, this entrepreneurial experience taught me to balance technical excellence with business viability.

What I Do

🔍 RAG Engineering

Building production retrieval pipelines with measured quality. Systematic evaluation, reranking optimization, and honest reporting of what works and what doesn't.

🤖 Agent Architectures

Designing agent systems with intelligent routing, tool integration, and cost tracking. L0/L1/L2 classification to skip unnecessary LLM calls.

📊 Eval-Driven Development

Every feature measured: latency, recall, cost, accuracy. Building evaluation frameworks and keeping honest records of experimental results.

🚀 Full-Stack AI Products

End-to-end product development: FastAPI backends, Next.js frontends, Docker deployment, SSE streaming, and production monitoring.

🗄️ Data Processing & Governance

Structured data pipelines, ontology extraction, and knowledge graph construction. AI disclosure on health endpoints, prompt hash tracking, and transparent cost accounting.

⚙️ Scalability & Modularity

Service-oriented architecture with independent deployable components. Modular RAG services that can be consumed by multiple products without tight coupling.

Technical Skills

Programming Languages

Python, TypeScript, Java, SQL

AI / ML

RAG, LLMs, Agent Systems, Knowledge Graphs, NLP, Prompt Engineering, Model Distillation (ongoing), HPC Training & Deployment

Frameworks & Libraries

FastAPI, Next.js, React, PyTorch, LangGraph, Autogen

Databases & Knowledge Systems

PostgreSQL, Neo4j, pgvector, Chroma

Infrastructure & Cloud

Docker, Docker Compose, AWS (via Udacity certifications), HPC clusters, Git, Linux

Certifications & Continuous Learning

Completed Udacity Nanodegrees and courses to stay at the forefront of AI engineering:

Nanodegrees (10)

• AWS Machine Learning Engineer
• AWS Cloud Architect
• Cloud DevOps Engineer
• Machine Learning DevOps Engineer
• Machine Learning Model Optimization
• AI-Powered Software Engineer
• Agentic AI
• Generative AI
• Responsible AI
• Agentic AI Engineer with LangChain & LangGraph

Courses (5)

• Agentic AI Fluency
• Building Generative Models
• DevOps Fluency
• MCP in Action
• Fine-Tuning AI Agents with Reinforcement Learning

Side Projects & Tools

Beyond core projects, I build tools to improve productivity and workflow:

📚 Chat Session Manager

Process downloaded chat session ZIPs into locally readable, queryable format. Supports embedding-based search and export to IntelliScribe for knowledge integration.

⚡ HPC CLI

Command-line tool for streamlined HPC cluster operations and job management.

📝 Quick Note Tool

Uninterrupted note-taking tool with video screenshots, text copy-paste, and AI Q&A capabilities (powered by llm-workflow modules).

AI Tools & Continuous Growth

I actively use and evaluate AI coding assistants to accelerate development:

Claude
Primary agent
Cursor
IDE integration
Qwen Code
Exploring
DeepSeek Harness
Exploring

I'm passionate about leveraging AI tools to grow with the field, while maintaining focus on experience accumulation — this combination helps me deliver products faster and with higher quality.

What I'm Looking For

I'm open to roles where I can:

  • Build production RAG systems and agent architectures at scale
  • Work on end-to-end GenAI applications — from experiment to deployment
  • Collaborate with teams that value measurement, honesty, and iteration
  • Continue growing in cloud infrastructure and enterprise AI deployment

Target roles: AI Engineer, RAG/Agent Specialist, Full-Stack + AI Engineer

Get in Touch

If my background and interests align with your team, I'd be delighted to discuss potential opportunities.