Home Blog Projects Tags About Friends
Z

Data + AI + Systems

Ziven Liu

I'm currently learning and building at the intersection of data engineering, AI-native systems, and system design, gradually shaping my own way of understanding complex systems.

What keeps me interested is not only how a tool is used, but why it exists, where it belongs in a system, and whether it genuinely helps solve the problem at hand.

Data Engineer AI Engineer AI Native Builder System Design

If I had to summarize it in one sentence, I'd say I'm trying to gradually place data systems, distributed foundations, and AI capabilities into one coherent engineering view.

Data Systems 95% AI Engineering 84% Distributed 82% Cloud Infra 86% System Design 90% Product Sense 72%

Identity

Capability Model

I prefer to think of myself as an engineer still growing in multiple directions: data engineering is the backbone, system design gives me structure, and AI engineering keeps opening up new possibilities.

Data Engineer

Build reusable systems around pipelines, layered modeling, semantic definitions, and platform capabilities.

AI Engineer

Move beyond isolated experiments and wire features, retrieval, and inference into production workflows.

AI Native Builder

Explore how AI can reshape data modeling, quality diagnostics, and data interaction experiences.

System-Oriented Engineer

Care more about why components exist, how boundaries are drawn, and how systems evolve over time.

From Data To AI

System Graph

To me, this graph feels more like a learning map. Some parts come from hands-on practice, some are still areas I'm actively studying, but they are gradually connecting into one network.

Data Systems

These days I naturally think in terms of pipelines, modeling, and semantics, and I'm trying to make data systems clearer, steadier, and easier to reuse over time.

Postgres Glue Iceberg Athena Trino

Distributed Foundations

Rather than only remembering how to configure things, I want to gradually understand the mechanisms and boundaries behind Kafka, Flink, Raft, and RocksDB.

Kafka Flink Raft RocksDB Consistency

AI-Native Infrastructure

I'm also learning how to connect RAG, recommendation systems, feature engineering, and Ray into real data platforms instead of leaving them at the demo stage.

RAG RecSys Feature Eng Ray LLM

Drag the nodes to see how these capabilities connect inside one system.

2019 2023 2024 2025 2026 Shanghai University of Finance and Economics Graduate Recommendation ByteDance Zulution VAST · Tripo

Trajectory

Evolution Path

This line is simply a record of a few stages that feel clear to me so far. There wasn't a dramatic leap, just a gradual push from study, internships, and work toward the intersection of data systems and AI.

2019

Shanghai University of Finance and Economics

Data Science & Big Data Technology

2023

Graduate Recommendation

Financial Information Engineering

2024

ByteDance

Data Platform & Warehouse Practice

2025

Zulution

AI Data Engineering

2026

VAST · Tripo

AI Data Infrastructure

How I Work

Structured Thinking

Problems are naturally decomposed into layers, modules, and stages before execution.

First Principles

I tend to ask for the underlying reason behind protocols, storage, and consistency instead of memorizing conclusions.

System Design First

I care more about whether a whole system is scalable, observable, and maintainable than about isolated features.

Future-Oriented

I keep thinking about how AI will reshape next-generation data architecture instead of only optimizing local concerns.

Engineering + Product

When designing solutions, I think about system boundaries, user experience, and how ideas are communicated.

Tooling as Leverage

I use AI coding, CLI workflows, and automation to squeeze out repetitive work and spend more time on architecture.

Find Me

If you're also thinking about data, AI, and system design, let's talk.