Customer workflow fluency · Agent-assisted delivery · Production workflows
Finance and audit trained me to read messy operations.Claude Code and modern engineering tools help me ship them as prototypes, tools, and production workflows.
02 / About
I'm Rain Fan, finance- and audit-trained. Three years at KPMG, workflow automation at SF Tech, and operating analysis in a startup taught me to enter a business context and understand how people, systems, spreadsheets, and exceptions actually work together.
Since 2024 I have been bringing AI into finance and operations workflows. By 2026, I can use Cursor, Claude Code, Python, and modern web tools to turn real business requirements into working prototypes, internal tools, and production systems.
I'm not positioning myself as a traditional frontend-only engineer. My edge is reading workflows, defining boundaries, and collaborating with agents to ship systems that actually run.
03 / Experience
PRODUCTION CUSTOMER SYSTEM
A full digital ops stack for a Minsk-local fried chicken shop
5 wks
shipped
1st
paid order
Key insight
The most counter-intuitive lesson from shipping full-stack with agents: the stronger the agent, the more important it is to plan first. Define the architecture, the boundaries, what the agent can and cannot touch—before writing the first line of code. Otherwise the implementation cost AI saves you gets eaten by the cost of going in the wrong direction.
INTERNAL WORKFLOW AUTOMATION
Monthly KA reporting + acceptance flow automation
6 reports
per month
1 day → 2 hrs
4× faster
Key insight
The starting point for AI automation isn't the technology—it's the pain. You have to have been repeatedly tortured by a process before you can really think through "why has it always been done this way," and only then can you tell what should be automated and what shouldn't. The prerequisite for building the right tool is real pain.
DATA OPERATIONS AUTOMATION
Multi-platform revenue reconciliation automation
+30%
efficiency
3 platforms
iOS / GG / PP
Key insight
In scaled finance work, the real lever isn't doing each thing faster—it's abstracting the shared rules into a model. Build it once, use it 100 times.
BUSINESS PROCESS DIAGNOSTICS
Listed-company audits and financial DD
30+
subsidiaries consolidated
100+
major contracts reviewed
Key insight
To deeply understand a company, you start with the financial statements—but the real insight is in the business flows of sales, procurement, production, and cash.
04 / Contact
I'm looking for AI implementation, Forward Deployed Engineering, AI Solutions, and Agent Workflow Automation roles. Preference: Shenzhen or remote.