APPLIED AI & PLATFORM ENGINEERING

Human curiosity.
Real-world AI.

Hi, I’m Rick. I build useful AI systems, and care about the people who use them.

At RWTH IT Center, I work on KI:connect: model APIs, backend engineering and the details that make an AI platform work in practice.

Aachen, Germany Chinese roots. An international outlook.

Ruizhang Zhou
CURRENTLY BUILDINGKI:connect · RWTH IT Center
Engineer. Always a learner.
Engineering AI platformsKI:connect / RWTH IT Center
M.Sc. Computer ScienceRWTH Aachen University
From languages to systemsShanghai → Aachen

01 / SELECTED WORK

Ideas are a start.
Making them useful is the work.

Real projects, clear contributions, and a few of the decisions behind them.

PERSONAL PROJECT · COMMUNITY FORK

TalkEcho

Meeting captions and dictation for a multilingual working day. My changes focus on Windows workflows, request resilience and provider-key handling.

TypeScriptTauri / RustSpeech-to-text

INDEPENDENT PROJECT · OPEN SOURCE

LLM Gateway

Choose the job, not a hard-coded model. Task-aware routing with model discovery, provider fallback and a deliberate boundary around streaming retries.

PythonModel routingReliability

MASTER’S THESIS · RWTH AACHEN

Learning to generate
traffic scenarios.

TimeGAN, diffusion models and real-world trajectory datasets for research into autonomous-vehicle testing. The research foundation behind my applied AI work.

PyTorchTimeGANDiffusion-TS

02 / HOW I WORK

Care is in the details.

Good engineering is a sequence of decisions you can explain, test and stand behind.

01.

Start with the real problem.

What does someone need to do? What gets in their way? I want the answer before choosing the model, framework or abstraction.

02.

Make failure understandable.

A timeout, a broken stream and a wrong usage count are different problems. I trace the path, reproduce what I can and keep assumptions visible.

03.

Leave something others can use.

Readable code, clear explanations and honest boundaries matter. I care about the person using the software and the person maintaining it next.

03 / THE PERSON BEHIND THE WORK

Languages first.
Then code.
Always curiosity.

A route from Shanghai to Aachen, and from understanding language to building systems.

LanguagesForecastingDivingLearning by building

My route into engineering started with languages. I studied German language and literature at Tongji University before studying computer science at RWTH Aachen.

That background still shapes what interests me: how people understand things, where communication breaks down and how a useful tool can close the gap. AI brings those questions into my daily engineering work.

Outside work, I’m interested in forecasting, learning languages and diving. Different ways to stay curious, think carefully and pay attention.

2024 — now
RWTH IT Center

Software engineering · KI:connect AI platform

2022 — 2024
RWTH Aachen University

M.Sc. Computer Science

2019 — 2022
RWTH Aachen University

B.Sc. Computer Science

2014 — 2018
Tongji University

B.A. German Language and Literature

A SMALL ENGINEERING PLAYGROUND

One choice.
A different experience.

A tiny example of the decisions behind the projects.

TRY THE ROUTING LOGIC

When should an AI request switch providers?

Before an answer starts, a fallback can help. After the first text appears, silently switching models could mix two different answers.

Interactive illustration. No API requests or personal data.

01Primary provider
02Next provider

Ready for an example.

See the actual routing policy ↗

04 / FROM THE NOTEBOOK

Things worth
thinking through.

Browse the full notebook ↗

A GOOD CONVERSATION IS A START

What are you
trying to build?

If your team is working through a difficult AI engineering problem, I’d be glad to hear about it.

ruizhang.zhou@mail.com