AI + Finance
Building AI applications for financial analysis, research storytelling, and narrated videos, with a focus on data definitions, content generation, and result validation.
Financial data storytelling ↗
LEO ZHOU / CHANGJU ZHOU
Data Visualization Lead · Hithink RoyalFlush AI
I lead data visualization in the AI department at Hithink RoyalFlush, with a background in designing and developing financial software and visualization systems. I currently focus on AI applications for finance and visualization, advancing a data visualization platform and iterating on an autonomous development workflow based on an agent loop in actual project development.
WRITING AS DATA
298 ESSAYS01 / FOCUS
Building AI applications for financial analysis, research storytelling, and narrated videos, with a focus on data definitions, content generation, and result validation.
Financial data storytelling ↗Building on HiVis and Standard Chart, I develop chart components and generation protocols for agents. Models produce chart specifications; code validates and renders them.
Visualization Agent ↗02 / SELECTED WORK

01 / DataV · HiVis · Standard Chart
I designed the core architecture of HiVis and led project and technical design for DataV and Standard Chart, while contributing to core features including component interfaces, cross-framework integration, and version evolution.
Standard Chart Sankey output

02 / Narrative Visualization
I am responsible for the overall project and technical design, and contributed to some early core development.
Whiteboard narrative video: cover of a generated financial explainer

03 / Viz Platform
I design the platform architecture, protocols, task definitions, and acceptance criteria, and built an autonomous AI workflow based on an agent loop.
AI-generated concept illustration: platform callers and chart output
03 / WRITING
01 / 2026-05-28
How AI changes engineering work from writing every line to designing context, tools, constraints, and evaluation loops.
3 MIN READ02 / 2026-05-18
A practical method for turning vague improvement goals into measurable feedback loops for both people and AI agents.
3 MIN READ03 / 2026-05-25
Why reliable AI-assisted engineering depends on observable feedback loops, executable checks, and continuously improved context.
3 MIN READSTAY IN THE LOOP
Subscribe in your RSS reader. No email address needed.
04 / RESEARCH & RECOGNITION
05 / NOW
Build something meaningful