WEEK 01

WEDNESDAY 09 / 02 / 26 · 6:30 – 9:30 PM

Introduction

Students leave knowing what the course is, what they will be expected to produce, and with every account and tool already set up — so Week 2 can start with code instead of installers.

Deliverables

SESSION PLAN

Four parts

PART 1

Course Introduction

Introduce ourselves, walk the syllabus, and say what “AI prototyping” actually means — where AI-assisted building is genuinely good, and where it is confidently wrong.

PART 2

Expectations & Policies

The midterm, the final, and the weekly deliverable. Department policies on attendance and grading. The AI use policy: the tools are the subject matter, but you own the output.

PART 3

Toolchain Overview

Name each piece and what job it does. Not how to use it yet — just why it exists.

PART 4

Account Setup

Every account created and working before you leave the room.

PART 1

Course Introduction

Who we are, what the semester looks like, and what “AI prototyping” actually means — including the places it is genuinely good and the places it is confidently wrong.

Who is teaching this

Two working engineers, not full-time faculty — which is the point. Everything in this course is a tool one of us used at work this month.

PHOTO

David Mendez

Staff Software Engineer · CNN

New York City

A full-stack engineer with over a decade of experience building dynamic, data-driven applications — educational games, agency work, and now real-time applications and interactive maps that power live news coverage and editorial storytelling.

Built an Electron desktop app for live election data visualisation — real-time Node.js/Socket.io server, React Redux Observables, drawing tools and draggable pop-outs.
Led architecture on a white-label React Native app that let a broadcast network ship 30+ station apps from a single codebase.
Guest lecturer at FIT since 2023 — AI techniques and no-code platforms for the Brand Experience Design and UX/UI Master programs.

BFA, Computer Animation and Interactive Media — FIT

linkedin.com/in/dave11mj dave.digital

PHOTO

Hedy Deng

Senior Software Engineer · WW International

New York City

A senior software engineer with over a decade of experience building performant, scalable and accessible web applications — from government sites to media platforms serving millions of users. Full-stack mindset, and a bias toward clean code and thoughtful UX.

Core developer on the ww.com migration off Acquia Cloud onto WW’s internal AWS infrastructure, and the legacy React class-to-function rewrite in TypeScript.
Works across React, Next.js, TypeScript, GraphQL, Drupal, AWS CDK, Docker and Kubernetes — plus Claude Code, Cursor and Copilot day to day.
Adjunct Professor at FIT and faculty at the Design Portfolio Center, mentoring designers on portfolio structure and performance.

BS — Cornell University

linkedin.com/in/hedy-deng hedyd.dev github.com/hedyd

Walking the syllabus

Fifteen weeks, laid out below. A midterm briefed in Week 8, built in Week 9 and due in Week 10; a final you propose yourself and present in Week 15; and a weekly deliverable that is always the same shape — something pushed to a repository.

WEEK

DATE

TOPIC

WHAT HAPPENS

Sep 2

Introduction

Course intro, expectations and policies, the toolchain, and every account working before you leave

Sep 9

Your First Webpage & Git Basics

Hand-code a page, submit it through Classroom 50, and publish it to a live URL

Sep 16

Enter AI Coding

A real development environment, and prompting as a skill — ship an opinionated page on AI ethics

Sep 23

Static App I — Stack & Scaffolding

Claude as a technical advisor; scaffold a statically generated app and write DESIGN.md

Sep 30

Static App II — Compiling & Deploying

Studio build of a pre-existing design, then a live side-by-side review

Oct 7

Full Stack App I — Wireframing

Systematised desktop and mobile wireframes with Cowork, deployed to Vercel

Oct 14

Full Stack App II — AI Inside the Product

Calling a model API from your own application

Oct 21

Animations & Interactivity

3D scenes with React Three Fiber and animation with Motion. Midterm brief released.

Oct 28

Midterm Work Session

A full session dedicated to building the midterm

Nov 4

Sharpening Your Tools & Accessibility

Midterm due. Custom Claude commands and skills, plus an accessibility pass on your own project.

Nov 11

Authenticated App I — Databases

Databases in plain language; modelling your data with accounts in mind

Nov 18

Authenticated App II — Production Readiness

The work between “it runs” and “it ships”. Final project announced.

Nov 25

Final Project Work Session

A full session dedicated to building the final project

Dec 2

Final Presentation Review

Work-in-progress presentations, structured peer critique, and final scope decisions

Dec 9

Final Presentations

Each student demos their live, deployed application

What “AI prototyping” actually means

Using an agent to get from an idea to something running, fast enough that throwing it away and trying again costs you an evening rather than a month. The bottleneck moves from typing to deciding. What it replaces is the blank page and the hours of boilerplate — not the judgment about what is worth building.

Genuinely good at

Scaffolding a project from nothing — config, folder structure, boilerplate
Turning a clear description into working code in a language you only half know
Renaming and refactoring across dozens of files at once
Explaining unfamiliar code you have inherited, line by line

Confidently wrong about

Library APIs that changed after its training data was collected
Whether the thing it just wrote actually runs
Your project’s conventions — unless you have written them down
Saying “I do not know” — it will invent something plausible instead

PART 2

Expectations & Policies

What you will produce, how the 100 points break down, and the one policy that matters most in a course whose subject matter is the tools themselves.

Midterm Project

20 points

A written brief lands in Week 8. Interpreting it, deciding what to build, and designing it are part of the assignment. Week 9 is a full in-class build session; the project is due in Week 10.

Final Project

30 points

You propose and build an original application of your own choosing. It must be deployed, and it must use a real data layer or an AI-powered feature. You own the scope — our job is to make sure the scope is honest.

Weekly Deliverables

30 points

Accept each assignment and push your work before the deadline. Pushing is submitting; there is no separate upload step. Every repo needs a README listing what you built and the live URL.

Professionalism

20 points

Attendance and punctuality, active participation, presentation quality, and willingness to debug, ask questions, and help the person next to you.

Attendance — the CT&D department policy

Class meets Wednesdays, 6:30 – 9:30 PM. You are expected to attend all classes, be on time, and stay for the whole session.
Miss three classes and the grade is an F. There is no way around this one.
Arriving more than 10 minutes after the start counts as late. Two lates equal one absence.
Arriving over 30 minutes late, or not returning from the break, is recorded as an absence.
An excused absence is still an absence — the difference is that it does not hit your professionalism grade.
If you are going to miss a class, email ahead. Class time will not be spent re-running a missed lecture.

AI USE POLICY

The tools are the subject matter. You own the output.

Use Claude, Copilot, Cursor — whatever gets you moving. But the work you hand in is yours. You must be able to explain any line of it, and over the semester you will be asked to.

PART 3

Toolchain Overview

Name each piece and what job it does. Not how to use it yet — just why it exists, and when you will first touch it.

TOOL

WHAT IT IS FOR

GitHub

Where the code lives and how versions are tracked

Classroom 50

Where assignments are handed out, submitted, and returned with feedback

VS Code

The editor everything happens in

Claude Code

The agent that writes and edits the code

Figma

Where the design system is defined

Vercel

Where the finished apps get deployed

Classroom 50, so it is not mysterious

It has no database of its own — it is a thin layer over GitHub. An assignment is a repository, submitting is pushing to it, and feedback is comments on a pull request.

Accept link

the instructor posts it

Your repo

in fitnyc-ct390

git push

that is the submission

Feedback PR

where comments land

Everything you learn about Git this semester is the same skill as turning in homework.

PART 4

Account Setup

Every account created and working before you leave the room. Five accounts — and the second one has an order to it, because nothing in Classroom 50 is visible until you are inside the GitHub organisation.

GitHub account

Create the account, then apply for GitHub Education with your FIT email. Approval is not instant, so start it tonight.

github.com/signup

Applying for GitHub Education, screen by screen

1Visit github.com/education for an overview of the program.

2Make sure your GitHub account is tied to your school email.

3Go to github.com/settings/education/benefits to start the application.

4Choose Student, and share your location when the pop-up asks.

5Upload a school ID — you should see “Your application has been submitted”.

6The application shows as pending until it is approved or rejected.

7Rejections usually mean the billing name does not match your ID.

8Fix whatever was flagged, then simply re-apply.

9Once approved, wait up to 72 hours before the benefits switch on.

Classroom 50 account

Two things, in this order. Classroom 50 is where every assignment is handed out and submitted — but it cannot see you until you are inside the class GitHub organisation.

FIRST

Accept the fitnyc-ct390 invitation waiting in your inbox. Nothing in Classroom 50 is visible until you are a member of the organisation.

check your email

THEN

Sign in at classroom50.org with GitHub — not with an email and password. Confirm that fitnyc-ct390 appears with a Student label beside it. If it does not, the invitation has not gone through yet.

classroom50.org

What you should see

SCREENSHOT · 1600 × 900

What the fitnyc-ct390 organisation invite looks like in an inbox, and where the Accept button is.

SCREENSHOT · 1600 × 900

Classroom 50 after signing in with GitHub — fitnyc-ct390 listed with a Student label beside it.

Figma account

Apply for Figma Education with your school email — an education seat is what lets you be part of an educational team.

figma.com

Claude account

Pro or above. The free tier runs out of usage partway through a single build session, which is exactly the wrong moment for it to stop.

claude.ai

Vercel account

vercel.com

While you are in there

GitHub Education students get Copilot Pro at no cost. This course centres on Claude Code, but the application takes a minute and the seat is free — worth doing while the GitHub tab is already open.

DELIVERABLES

Every account working before you leave

GitHub account created, with GitHub Education applied for

The email invitation to the fitnyc-ct390 organization accepted

Signed in at classroom50.org with GitHub, showing fitnyc-ct390 with a Student label

Figma account, with Figma Education applied for

Claude account, Pro or above

Vercel account