FORWARD DEPLOYED ENGINEER · AI DELIVERY

Make AI Actually Run in Your Business

18 years of hands-on delivery across Huawei, Alibaba, and Tencent ecosystems. We don't pitch concepts — we come to your business on-site: a 2-week sprint to find the AI use cases that actually make money, then 6 weeks to get them live.

Others end with a slide deck. We end with a live system.
🏛 In support of MIIT's Special Action for Cultivating AI Application Service Providers (Doc. No. 414 [2026], General Office of MIIT)
The policy explicitly encourages "building forward deployed engineer (FDE) teams that stay on-site with customers" — exactly what we've been doing for 18 years
Book a Free Discovery Call See Our Services
18+Years of ICT Delivery
¥10M+Presales Wins Supported
40+Data & AI Projects Delivered
30+Corporate Training Sessions
WHAT IS FDE

What Is the FDE Delivery Model

FDE (Forward Deployed Engineer) was pioneered by Palantir and is now one of the most contested roles at top AI companies like OpenAI — people who understand the technology and the business, embed directly at the customer's site, and get AI running inside real business processes.

In the AI Era, the Scarce Thing Isn't Models — It's Delivery

Companies can buy tools, attend workshops, and collect slide decks — yet they get stuck on the last mile:no one can embed AI into real business operations and make it run. Model capability is abundant; the ability to deliver is not.

Yuexiang Tech turns FDE into a service you can buy: a team with 18 years of delivery experience embeds at your site — first maps the real workflows, then sizes up the ROI, and finally gets it live in production.

"Our engineers sit next to your data."
— the original definition of the FDE model, and the way we deliver
Dimension
Training Vendors
Traditional Consulting
Yuexiang FDE Delivery
End Point
Course Ends
Report Delivered
Live in Production
Who Does the Work
Instructors
Consultants
Delivery Engineers On-site
ROI Focus
No ROI Talk
Mostly Frameworks
ROI First, Then Build
Data Foundation
Not Covered
Sometimes
Data Governance Is Our Baseline
What Not to Do
—
Rarely Said Outright
We Tell You Not to Do It
📌 Policy backing: MIIT's Special Action for Cultivating AI Application Service Providers (Doc. No. 414 [2026], General Office of MIIT) has written "forward deployed engineer (FDE) teams" into the supporting measures for cultivating service providers. That means FDE is no longer just our commercial pitch — it is a delivery model recognized by national policy. Choosing FDE means moving in the same direction as the policy.
PRINCIPLE 01

ROI First

We calculate the return before we start. If a use case doesn't make financial sense, we tell you straight — we don't survive by selling big deals.

PRINCIPLE 02

On-site Delivery

We spend half a day working alongside your frontline staff, watching how work actually gets done — not listening to second-hand requirements.

PRINCIPLE 03

Data Foundation

90% of AI delivery problems come from data, not models. A data governance background is our core advantage.

SERVICE PRODUCTS

Three Product Lines. You Buy Outcomes.

Fixed price, fixed timeline, fixed deliverables — buy AI delivery like buying a product

🔍

AI Opportunity Sprint

2 weeksFrom ¥35,000▸ Consulting & Planning

A complete inventory of AI opportunities inside your company, delivered as a roadmap your leadership can approve on the spot: which use cases are worth doing, how much they save or earn, and which one to start with.

  • On-site research: executive interviews + frontline shadowing + data review
  • 5–10 candidate use cases, each vetted for technical feasibility
  • Use-case-by-use-case ROI estimates and priority ranking
  • Leadership presentation + a 90-day implementation roadmap
🚀

AI Use Case Launch

4–6 weeksFrom ¥80,000▸ Delivery & Implementation

We take one high-impact use case from idea to live production. No oversized new systems — built on mature platforms, so your business team can use it on day one.

  • Scope and acceptance criteria frozen (quantified, signed by both sides)
  • Data connected, access granted, MVP running end to end
  • Pilot with a real business team, daily iteration
  • User guide + ops manual + team training
🤝

AI Ongoing Support

MonthlyFrom ¥8,000/month▸ Operations & Management

Weekly on-site and remote support after launch: continuous tuning of live use cases, plus coaching to build your own AI delivery team — so the capability stays inside your company.

  • Continuous tuning and issue response for live use cases
  • Rolling discovery and piloting of new use cases
  • One-on-one mentoring of your internal staff
  • Monthly results review and improvement recommendations

Five Service Capability Categories

MIIT classifies AI application service providers into five capability categories. We cover all five — buy any single category, or combine them into a complete delivery chain.

① Consulting & Planning
AI Opportunity Sprint
Strategy Consulting · Use Case Insight · Solution Design
② Delivery & Implementation
AI Use Case Launch
Custom Development · System Integration · Platform Tooling
③ Operations & Management
AI Ongoing Support
System Ops · Iteration & Tuning · Team Mentoring
④ Security & Governance
AI Security & Compliance
Data Security · Risk Alerting · MLPS & Xinchuang
⑤ Supporting Services
AI Talent Coaching
DAMA Certification · Corporate Training · Assessment
🛡️

AI Security & Compliance

On-demand assessment▸ Security & Governance

We embed cybersecurity, data security, ethics, and regulatory compliance into the full AI lifecycle — from R&D to deployment to daily use. Proactive, not reactive — exactly what the Special Action requires of service providers.

  • Data security and access-control assessment for AI applications
  • MLPS / Xinchuang alignment and security hardening advice
  • AI risk alerting and compliance management systems
  • Anonymized deliverables and end-to-end client confidentiality
🎓

AI Talent Coaching

By cohort▸ Supporting Services

The policy calls for batch-training versatile talent that knows the business, the models, security, and delivery. We've turned 18 years of delivery experience into courses, and we also co-build training bases.

  • DAMA Certification Training (CDGA / CDGP / CDAM)
  • Custom corporate training: picking AI use cases from real workflows
  • On-site mentoring — the capability stays in your organization
  • Testing, assessment, and training-effectiveness metrics

Foundation Capabilities (18 Years in the Making)

Data Governance Consulting & Implementation Data Asset Inventory & Balance-Sheet Entry Data Platform Planning & Build-out DAMA Certification Training (CDGA / CDGP / CDAM) Custom Corporate Training Custom Software Development Project Management Systems & PMO Setup CMMI / ISO Certification Consulting
DELIVERY PROCESS

Delivery Process: Every Step Visible

A cadence built over 18 years of project delivery — what happens when, and what gets delivered, spelled out up front

AI Opportunity Sprint · 14 Days

From ¥35,000
D1

Kickoff

2 hours online: align on goals and scope, send the prep questionnaire and project charter

D2-D4

On-site Research

Executive interviews, frontline shadowing of real workflows, systems and data review

D5-D8

Use Case Mapping

5–10 use case cards, each vetted for technical feasibility and data readiness

D9-D11

ROI Calculation

Three-year ROI estimates per use case; ranked down to 3 P0 priorities

D12-D13

Leadership Presentation

Page 1 is the ROI summary; the last page is "use cases you should NOT do"

D14

Handover

Roadmap handed over, with options for what comes next

AI Use Case Launch · 6 Weeks

From ¥80,000
W0

Scope Freeze

Quantified acceptance criteria, signed — this week decides whether the project becomes a money pit

W1

Data & Environment

Data connected, accounts provisioned, system interfaces opened up

W2-W3

End-to-End MVP

Full use case running end to end, plus a 3-minute demo video

W4

Real-User Pilot

1–2 real business teams using it daily; issues collected and fixed daily

W5

Training & Acceptance

Results check, user and ops manuals, team training, acceptance sign-off

W6

Review & Next Steps

Closing report with measured ROI, and options for ongoing support or a second use case

Research Notes: Sent same day — no questions left overnight
AI Opportunity Map: the whole opportunity landscape on one page
Use Case Cards: pain point, solution, data, returns, risks
ROI Model: open formulas, adjustable parameters
P0 Use Case Proposal: project docs ready for approval
90-Day Roadmap: who does what, by when
SCENARIOS

Industries & Scenarios We Know Well

18 years deep in data and AI — we know the tech stack and the real-world numbers of every industry we serve

📦

Cross-border E-commerce & Trade

Store, ads, supply chain, and fulfillment data scattered across platforms; selection and inventory still run on gut feel.

Typical entry points: AI product selection, customer service automation, multi-platform data integration
🏭

Manufacturing & Industrial Clusters

Owners of apparel, beauty, and leather-goods clusters want to transform but don't know where to start.

Typical entry points: quality-inspection assistance, inventory & production forecasting, livestream analytics
⚡

Energy & Utilities

Massive volumes of equipment and metering data sitting idle; the value of data assets is untapped.

Typical entry points: data asset inventory & balance-sheet entry, data quality governance, report automation
🏛️

Government & SOEs

Data scattered across systems and departments — hard to share, inconsistent definitions, reports rebuilt again and again.

Typical entry points: data cataloging & sharing, unified metric definitions, intelligent Q&A and document drafting
🚗

Automotive & Mobility

Huge volumes of vehicle and business data, with high compliance and governance requirements.

Typical entry points: data governance framework, document & process automation
🏢

Professional Services & SMBs

Small teams, heavy workloads — repetitive documents and knowledge searches eat the day.

Typical entry points: knowledge base Q&A, contract & proposal first drafts, automatic ticket triage
Policy industry alignment: these scenarios cover the major service industries named in the Special Action — manufacturing · energy · commerce · retail · government · transportation · software · smart city. For each, we deliver the full chain from diagnosis to implementation to ongoing support.
DELIVERED

Delivered Work

From government and energy to SMB systems — 18 years of actually shipping these things

⚡

Energy Industry Data Lake

Data lake build-out and cluster optimization

EnergyData Assets
Delivered
📋

Data Asset Planning

Planning and program design for data assets

GovernmentBalance-Sheet Entry
Delivered
🏛️

Provincial Government Platform

Program management and launch of a government app

GovernmentSmart City
Delivered
🧭

Data Governance Consulting

Governance standards and asset inventory

EnergyGovernance
Delivered
📊

Data Platform Build-out

DW architecture and BI reporting

ManufacturingData Platform
Delivered
🛒

E-commerce Group-Buy Platform

Full-featured e-commerce platform

RetailCommerce
Delivered
💰

Payroll System

HR digitization

SoftwareCommerce
Delivered
📅

Meeting Management System

Enterprise meeting digitization

SoftwareCommerce
Delivered
🏢

Luxury Property Mini-Program

Property services for One Shenzhen Bay

Smart CitySmart Home
Delivered
📚

Education Platform

Online learning and course management

EducationSoftware
Delivered
🏠

Smart Home App

IoT device orchestration

Smart HomeTelecom
Delivered
🔐

Secure Access System

Enterprise network and access security

CybersecurityPublic Safety
Delivered
Government, energy, and finance projects are under client confidentiality — client names and data details are not published. Happy to walk you through comparable delivery experience on a call.
AI APPLICATION SERVICE TEAM

Leading an AI Application Service Consortium

The Special Action requires service providers to lead consortia — one lead provider with at least two upstream and downstream partners. We're forming ours now, and you're welcome to be part of it.

Why a Consortium — Not Going It Alone

Real enterprise AI delivery is never a one-company job: where does the compute come from, which model to choose, how to fix the data, how to wire up the systems, who runs on-site — miss one link and you're stuck at the last mile.

Our role is lead provider + on-site delivery team: our FDE team works at the business site to define use cases and acceptance criteria, then assembles capabilities from compute, model, data, and device partners into one integrated solution.

🧠
LLM / Agent Vendors

They supply models and tokens; we package the use cases and deliver

⚙️
Compute & Cloud Providers

We tap compute vouchers and platforms to cut your model costs

🗄️
Data Providers

Data supply, governance, and balance-sheet entry, end to end

📱
Device & Application Vendors

Hardware/software adaptation, multi-model orchestration, and systems integration

What You Get as a Consortium Member

  • A real path to production for your technology — not another stalled POC
  • Requirement clarification and quantified acceptance criteria from our on-site FDE team
  • Jointly built, reusable "small, fast, light, precise" solution packages
  • Joint submissions of flagship solutions and consortium results — shared policy visibility
  • Access to supply-demand matching, first-purchase programs, and overseas expansion services
We also help fellow service providers prepare resource pool applications: AI case write-ups, service handbooks, credentials and supporting documents. If interested, pick the matching service type in the form at the bottom of this page.
SELF-DEVELOPED PRODUCTS

Products We Built Ourselves

Tools forged from 18 years of delivery — we built them to solve our own project problems first, now they're open to you

Yuexiang DataIn

Yuexiang DataIn

V1.0 Public Beta Software Copyright Pending

A lightweight tool that loads Excel / CSV into MySQL in one click: drag in a file and it detects column types, generates the DDL, cleans and validates, then writes straight to your database. All processing happens locally in your browser — nothing uploaded, nothing retained. Enterprise data security by design.

AI type inference Auto-generated DDL Cleaning & quality reports Reusable data templates Local processing · data never leaves your machine Free with registration
Use It Online Now Backup Link Online · 3 free tries/day, no signup
Yuexiang FDE Delivery System

Yuexiang FDE Delivery System

V1.0 Live Demo Built from field practice

18 years of delivery experience, codified into a system: project ledger, Gantt progress, delivery SOPs, deliverable templates, capability framework, knowledge base, case library, and cost estimation — all in one backend. Every project runs on the same standard, and every finished project becomes an asset — no more living in people's heads and laptops.

Delivery overview dashboard Project ledger & Gantt charts Delivery process SOPs Deliverable & template library Capability framework · Knowledge base Reusable case library Quoting & cost estimation
Try the Demo Backup Link Front-end demo · data stays in your browser
TEAM

Core Team

From Huawei and other top firms — 18 years in the industry, writing code and leading projects

Mr. Xu

Mr. Xu

Founder & CEO · Head of AI Delivery (FDE)

18 years in ICT, from developer and requirements analyst to project manager, program manager, and delivery director. 5.5 years in Huawei software; presales architect/delivery director for an Alibaba-backed data intelligence product line; big data presales architect at a Tencent partner. Lifetime member and certified instructor at DAMA China, holding 9 professional certifications including CDGA/CDGP/CDAM/CCDO. Supported presales wins of ¥10M+, delivered 40+ data and AI projects, and led 30+ corporate training sessions.

AI DeliveryData GovernanceCCDOPMPCMMI
Mr. Li

Mr. Li

Product Director

7 years at Huawei, core developer on the NGBOSS system. Fast learner, strong sense of ownership, consistently high delivery quality.

Java DevelopmentPMPTeam Management
Mr. Zhou

Mr. Zhou

Operations Director

Deep internet veteran, quick self-learner and creative thinker — strong in product design, requirements analysis, and data analysis.

Product DesignData AnalysisOps Planning
Mr. Zheng

Mr. Zheng

Technical Expert

10 years at Huawei developing the NGBOSS product line. The go-to expert for thorny technical problems, always digging into new tech.

ArchitectureJava DevelopmentProblem Solving
Mr. Huang

Mr. Huang

Senior Developer

5 years at Huawei, core developer on the ESOP platform. Humble, curious, and strong in framework design and new-technology research.

Framework DesignJava DevelopmentNew Tech Research
Mr. Zhang

Mr. Zhang

Full-Stack Developer

Energetic and driven, loves sports and coding. Works with HTML, JS, and jQuery on the front end; SSH and SSM frameworks; MySQL and Oracle databases.

FrontendBackendDatabases
INTELLECTUAL PROPERTY

Intellectual Property

Multiple software copyrights and in-house products, driven by continuous innovation

ShareBook copyright certificate

ShareBook

Louba Online copyright certificate

Louba Online

QueueUp copyright certificate

QueueUp

CEO Academy copyright certificate

CEO Academy

Yuexiang Software Store copyright certificate

Yuexiang Software Store

FAQ

Questions You Might Have

Putting the hard questions on the table up front is a lot cheaper than arguing afterward

How much does AI delivery actually cost?

Start with the Opportunity Sprint: ¥35,000 (¥15,000 for the first batch of case clients) gets you a complete roadmap and ROI model — then decide whether to invest in implementation. The Use Case Launch starts at ¥80,000; you can validate value with a single use case first.

Our data is a mess — do we have to finish governance before doing AI?

No. During diagnosis we assess data readiness for every use case and prioritize the ones that can run on the data you already have. Data governance is a long game, but AI delivery can move in batches.

Will you make management decisions for us?

No. We make technical feasibility, costs, and returns clear, and put the options in front of your leadership. Whether to proceed and how much to invest — that call stays with you.

Can we maintain it ourselves after the project ends?

Yes. We build on mature platforms, not black-box custom code, and we hand over user guides, ops manuals, and training. For extra peace of mind, choose monthly ongoing support and we'll bring your own team up to speed.

How do you guarantee data security?

Client data never leaves the environment you designate. We sign NDAs before engagement and anonymize client information in deliverables. For government and energy projects, we're familiar with MLPS and Xinchuang requirements.

Are you a government-designated AI application service provider?

The Special Action is organized by the Ministry of Industry and Information Technology, with provincial authorities building the provider resource pool; first-round submissions are due by December 1, 2026. We're preparing our provider file, credentials, service handbook, and case evidence. We won't claim selection before authorities publish the list — but we're happy to show you our current credentials and compliance status.

We want to apply to the resource pool ourselves — how can you help?

Gladly. We offer AI case write-up and scoping, application prep and self-review, service handbook compilation (drawn from our FDE Delivery System SOPs), credentials checklists, and — crucially — real use case delivery. The first 30-minute call is free: we'll tell you whether applying is worth it.

What is the AI Application Service Consortium, and how do I join?

The Special Action requires service providers to lead consortia — in principle one lead provider plus at least two upstream and downstream partners (compute, data, devices, models), submitted together with the resource pool information. We're forming a consortium now. LLM, compute, data, and device vendors are welcome to contact us, as are fellow service providers for joint applications.

What filings are required for generative AI?

If your service involves generative AI with public-opinion or social-mobilization capabilities, you must complete the Cyberspace Administration's generative AI service filing or registration; public-facing websites and mini-programs also need ICP filing and public-security network filing beforehand. During the solution phase we check whether your use case triggers these requirements — so you don't get shut down after launch.

CONTACT

Let's Talk for 30 Minutes — Free

No upfront cost. We'll help you decide first: whether your business has AI use cases worth investing in

Free Discovery Call

One 30-minute call: you describe your business and pain points, we give an initial read — which 1–2 use cases are worth digging into, and roughly how much they could save. If there's value, we continue; if not, we tell you straight.

📧

Email

rxid@163.com
londarxid@gmail.com

📞

Phone

+86 158 8963 3210

💬

WeChat

WeChat ID: londarxid

WeChat QR code
📱

WhatsApp

londarxid@gmail.com

📘

Facebook

londarxid@gmail.com

✖️

X (Twitter)

londarxid@gmail.com

📍

Service Area

Based in Guangzhou · Serving the Greater Bay Area & Nationwide