Fang Hao Xin Qing logoFang Hao Xin Qing
About / AI-native

Enterprise AI-native transformation starts with one real business problem.

Zhihao Xu | Enterprise AI-native transformation consultant

I bring AI coding, business workflows and agent delivery into the same room—solving one concrete problem first, then turning what works into team capability.

FDE enablement and business-agent deliveryAI coding and business-engineering trainerFinance / ERP / product workflow practitioner
Owners: see real deliveryLearn FDE: see the method
Zhihao Xu, enterprise AI-native transformation consultant

AI Native / FDE

Turn business problems into tools the team owns

2 yrs

AI voyage coach

10+

Offline talks and trainings

100+

Founders and operators guided

4

Real delivery cases

Real delivery

Four real deliveries across agents, workflow efficiency and team training.

Each case answers three questions: what was the business context, what changed, and what is still needed to scale it.

A semiconductor materials company / recruiting agentPartner: 得否科技

A recruiting agent delivered in two weeks

In partnership with 得否科技, delivered a runnable recruiting agent for a semiconductor materials company, covering resume acquisition, screening and interviewer notifications across multiple roles.

One agent can now cover more than eight roles, screen over 200 resumes per day and notify interviewers about suitable candidates.

Delivery outcome

2-week delivery

8+ roles
200+ resumes / day
Workflow live

Group enterprise / bank reconciliation

Six weeks → 10 workdays

40+ entities and 200+ bank accounts live in 10 workdays

Focused on bank reconciliation and daily cash reporting across more than 40 entities and 200 bank accounts.

The project was estimated at roughly six weeks and went live in 10 workdays.

40+ entities200+ accountsDaily cash report

Product team / AI-native workflow

10 days → 3 days

A normal PRD cycle reduced from 10 workdays to 3

Introduced AI-native work across requirements, business analysis, competitive research, PRDs, acceptance criteria and task breakdowns.

The product team's normal PRD cycle fell from 10 workdays to 3. Further adoption in engineering exposed the need for fuller change design at the organizational level.

The efficiency result was validated; organization-wide adoption needs alignment, sponsorship and a pilot mechanism.

Product workflowAI codingAdoption review

Wanxiang AI Lab / training delivery

100+ registrations

100+ registrations and 4 live sessions over two weeks

Co-delivered an AI content-acquisition training project with Wanxiang AI Lab, covering course organization, enrollment and online delivery.

The program ran for two weeks with four live sessions and a paid training engagement. It demonstrates training delivery, not a claimed business outcome for learners.

2-week course4 live sessionsPaid delivery
The FDE method

From a frontline problem to a tool the team owns.

FDE (forward-deployed engineering) is not just writing code on site. It connects business context, a working solution and reusable organizational capability.

01

Find the first pain

Start with high-frequency, repetitive work that crosses teams and can be measured.

Output: scope, workflow map and metrics

02

Let operators build

Train the people closest to the workflow instead of turning AI into a black-box demo.

Output: a first agent or business tool

03

Launch, measure, reuse

Track time, quality, adoption and reuse so each delivery makes the next one faster.

Output: usage data, SOP and review notes

AI voyage track record

Two years bringing AI coding into real practice arenas.

From AI coding, agents and MCP to AI product export, I have worked across Shengcaiyoushu voyages and brought the methods back into enterprise work.

AI Agent Architect

Use Codex, Hermes and Feishu CLI to turn one repetitive business task into a runnable, deliverable agent.

Voyage scale: 6,274 participants

Super AI Voyage

Build five AI capabilities in 14 days: thinking, accumulation, expression, product and execution.

Voyage scale: 14,729 participants

AI Coding Foundations (OpenClaw / Skills)

Start with Claude Code, Skills and OpenClaw to build AI coding, automation and always-on assistant workflows.

Voyage scale: 2,634 participants

AI Product Export (3rd edition)

Worked on the third-edition delivery around AI-coded products, overseas launch and sustainable operation, supporting the path from idea to validation.

Voyage scale: 1,500 participants · all three cohorts full

Claude Code in Practice

From environment setup to personal pages, small tools and lightweight business systems with Claude Code.

Voyage scale: 645 participants

Claude Code Programming Guide

Use Claude Code to build personal pages, small tools and Web applications from zero.

Voyage scale: 902 participants

MCP Applications

Connect AI to search, files, APIs and business tools through MCP configuration and workflows.

Voyage scale: 566 participants

DeepSeek Starter Guide

Use DeepSeek for content, design, video and office work, starting with fundamentals and prompting.

Voyage scale: 3,913 participants

Why me

From finance ERP to AI-native business engineering.

I have worked across enterprise delivery, product workflows, training and AI coding. My job is not to make AI sound abstract, but to place it inside the work an organization already does every day.

After graduating in 2016, I joined HAND as an ERP consultant and worked on enterprise finance systems and digitalization. I later spent years as a financial product manager, working across business, systems and organizational handoffs.

I worked across Shengcaiyoushu AI voyages for two consecutive years, delivered 10+ offline AI talks and trainings, and guided 100+ founders and operators in using AI coding for real business problems; from February to June 2026, I also held discovery conversations with 40+ companies through consulting.

Since 2026 I have been running a one-person company in public, turning delivery into reusable assets through the FDE Bluebook, Enterprise Finance AI Bluebook, AI coding manual and AI content systems.

Certified Public Accountant since 2019
Former ERP consultant
Two years as an AI voyage coach
Author of FDE and Finance AI bluebooks
Training format

Start with one real workflow and let the team build its first tool.

Training is not finished when the tools are explained. It is finished when the team can solve its own problem with AI.

WorkshopBuild around one real workflow
PracticeLet operators build the first tool
EnablementTrack adoption and reuse after launch
Training is not finished when the tools are explained; it is finished when the team builds its first tool.
What I can do

Start with one business problem.

For owners and teams that want training, workflow change and measurable results to connect instead of ending at an AI experience session.

Enterprise AI-native transformation

Select a measurable pilot from a high-frequency pain point, then design, adopt and review the workflow.

FDE and business-agent delivery

Deliver a working agent or business tool around finance, recruiting, product or operating workflows, with usage guidance and measurement.

AI coding and team training

Use workshops, practical tasks and enablement to help operators build their first tool.

How we work

Training is the start. Business outcomes are the finish.

I work best with owners, business leaders and education partners who are willing to drive change. We clarify the problem first, then decide whether a small pilot is worth doing.

Understand the business before designing the solution
Prove one critical workflow before scaling
Let operators build instead of only receiving training
Measure time, quality and adoption with real data
Start a conversation

Bring one real business problem.

Recruiting, bank reconciliation, product collaboration or content acquisition—we can first decide whether AI is worth using and where to start.

Zhihao's WeChat QR code

Add me on WeChat

Want to learn FDE? Start with the method assets.

Read the FDE Bluebook, then explore the finance AI and AI coding material to see the full path from frontline problem to delivery outcome.

FDE Bluebook: understand the role and delivery model
Finance AI Bluebook: see roles, workflows and governance
AI coding manual: turn an idea into a business tool
Read the FDE Bluebook