★ 第 25 单元

Ship and present

Your final project: one complete AI-assisted product of your own, from a real problem to a live link and a demo. Scope an MVP, plan which AI tools help at each step, design the core experience, build with AI coding tools, add and evaluate an AI feature, test, ship and present it honestly. Every lesson ends with a piece of your final product.

🔒 第 25 单元仅限会员

第 7 至 25 单元仅向训练营会员开放。如果你是学员请登录，或申请加入。

登录返回免费课程

1Choose and scope your final projectPick a real problem worth solving, cut it down to a minimum viable product, and write a one-page brief with success metrics you can actually measure.
2Plan an AI-assisted workflowDecide which AI tools help at each step, what stays human, and keep a decision and AI-use log you can build your case study from.
3Design the core experienceMap your core loop as a flow, choose the few key screens that carry it, set up a light design system, and assemble the key screen from components and variables.
4Build with AI coding toolsGenerate and iterate on code with AI assistants, review code you didn't write, use version control, and keep API keys on a server, never in the browser.
5Add the AI featureDesign the system around your AI feature: the system prompt, structured output, evals, guardrails, fallbacks and cost.
6Test and iterateRun quick usability tests and evals, fix the most important problems first, and do an accessibility pass with AI help and human judgment.
7Ship itDeploy your product to a real URL, run a launch checklist, and set up monitoring and feedback so you learn from real use.
8Present and pitchTell your final project's story as a case study and a live demo, and be honest about exactly what the AI did and didn't do.
★Your final project: an AI-assisted product, shipped and presented项目