Foundations → Applied → Enterprise
Three progressive tiers. You leave with SOPs and agents ready for production, and a full year of community support on every tier.
Deliverables you keep
One year of support
Accepted on results
The handover pipeline · bid first drafts
6 steps · 2 human checkpoints
Which tier fits you?
02 / Where you're stuck
Six months in, a handful of people still account for all the usage. Nobody told the rest what it's good for, or where the limits are.
The session went well. Back at their own reports, bids and customer queries, everyone stalls. The gap isn't concepts — it's turning a specific job into reusable prompts and workflows.
Budget and reporting are due, but which process goes first, how the return is calculated and who owns it afterwards are all unclear — so the proposal keeps slipping.
03 / Three tiers
Every tier includes a year of community support. The price on this page is the price you pay — no hidden fees.
Get everyone to a working baseline
¥3,960
per person · 2 days on site or 4 online sessions
For:Staff with no prior AI experience, admin and support functions
Covers:What current models can and cannot do, prompting fundamentals, rewriting everyday copy and documents, data and privacy red lines
You take away:20 prompt templates for your own role, plus a task reference sheet
Includes 1 year of community support
Turn each role's actual job into a workflow
¥7,960
per person · 3 days + one output review
For:Marketing, sales, operations, finance and HR — roles with a defined output
Covers:Breaking down role tasks, prompt engineering, multi-step workflows, output verification and rework loops
You take away:A prompt library for your role, 2 production-ready workflows and an SOP manual
Includes 1 year of community support
Process redesign and custom agents
From ¥136,000
per programme · 4–8 weeks · scoped by department size
For:Department or company-wide initiatives that need auditable results and material for reporting upward
Covers:Pre-course process diagnostic, teaching material written from your real data, agent and knowledge base co-development, management review
You take away:3 or more agents/workflows in production, an enterprise knowledge base and a signed acceptance checklist
Includes 1 year of a dedicated company channel
TIER 03 · Enterprise
We run a diagnostic on site before the course to find where cost comes out. Teaching material is written from your own business samples. By the time the course ends, the chosen process is already running.
Interview 3–5 roles, select 2–3 processes worth rebuilding, deliver a diagnostic note
Exercises use your real material; working workflows and agents get built in the room
Deliverables confirmed against the checklist, with a written result summary you can report internally
Covers onboarding support after staff rotate, plus a quarterly effectiveness review
03b / One year of support
Included on every tier, not just the expensive one. This field turns over every few months — being left alone afterwards is the real risk.
Instructor Q&A
Answered within 24 hours on business days, no cap on questions, two instructors on rotation
Monthly live review
One session a month on new tooling and real participant cases
Tooling and model updates
Prompt libraries and workflow files updated when major models change
Growing template library
Everything added during the year is yours to take
Peer network
Grouped by industry, for posting needs and finding partners
The boundary, stated plainly
Support covers tool usage and approach for course-related work. It does not cover delivering your projects for you.
04 / What you take away
Handed over with the rest of the files at the end. Fully exportable.
Prompt library
Grouped by role
Each entry states the role it suits, the input it expects and the output it produces. Ships with checklists and the rework points people hit most.
Why us
This section is for whoever has to explain the choice upward. Screenshot it straight into the deck.
—
Open course
Consultancy
Xingjie AI
We carry the risk
The checklist is signed by both sides before the course — not explained away afterwards.
If day one isn't worth it, you get your money back on the spot and keep the material.
A 60-minute live walkthrough on one of your real scenarios, before you commit to anything.
06 / Free capability assessment
You get a free diagnostic from the result — where you are now, which tier to start from, what to fix first. Three questions, about 30 seconds.
Nothing is collected while you answer. We only store and contact you if you choose to enter your name and phone on the result screen and tick consent.
08 / FAQ
是。价格含教材、Prompt 库、工作流文件与一年社群答疑,不再收资料费或平台费。异地内训的交通住宿按实际发生额报销,签合同前书面确认额度。企业档写「起」是因为按部门规模与流程改造范围计价,报价单会逐项列明。
入门档 2 天(可拆为 4 次线上,每次 2 小时);应用档 3 天 + 一次成果评审;内训档 4–8 周,其中集中授课通常 2–3 天,其余时间用于诊断、共建与验收。多数客户把授课排在周五与周六,或按班次分两批,避免同一岗位全员离岗。
签保密协议后再进入诊断。教材使用脱敏样本,敏感字段按替换表处理;如贵司要求,可全程在内网或私有部署环境操作,项目结束后删除我方留存的一切业务材料并出具确认函。课程同时会交付一份《AI 使用数据红线清单》,规定哪些内容不得输入外部模型。
范围是课程相关的工具使用与方案思路:提示词写不出效果、工作流卡在某一步、模型更新后文件要怎么改,都可以问,工作日 24 小时内响应、不限次数,固定两人轮值。不含代客交付具体项目——例如「帮我们把这份 80 页标书写完」属于项目外包,不在答疑范围内。
按成果验收。开课前双方确认验收清单,内训档至少交付 3 个正在运行的工作流或 Agent,附前后耗时对比;应用档每位学员带走本岗 Prompt 库与 2 条可投产工作流。清单未达成的部分,我们免费追加一次辅导。
可以。入门与应用档按人报名,先派一人参加再决定是否全员开班;内训档可安排 60 分钟内部试讲,用你们自己的一个真实场景现场演示,不收费。发票为增值税专用发票或普通发票,均可提供。
09 / Get in touch
We ask about your situation before quoting, and we don't mass-mail material. Enterprise clients can request sector-specific case material and a quote directly.

WeChat · alin499607
Xingjie AI · Enterprise AI Enablement · By YGG
YGG main site →