Course Syllabus
- Course name and ID: OIM3690-01, Web Development with AI
- Meeting time: Tuesday and Thursday, 3:00 PM to 4:30 PM ET, 9/01 to 12/08, 2026
- Location: Gerber Hall 102
Faculty Information
- Name: Dr. Zhi Li
- Office: Babson Hall 216D
- Email: zli@babson.edu
- Office hours: Thursdays, 4:45 PM to 5:45 PM, in person in Babson Hall 216D. Drop in, no appointment needed. For any other time, email me or use the booking link in our Canvas course.
Required 1:1 meeting. Every student completes one 15-minute meeting with me, in person, during the semester. It counts toward your participation grade. Book it whenever you are ready.
Course Description and Objectives
In this course you will build and deploy working web products with an AI agent, beginning in the first week.
Producing the code is now the least difficult part of that work. An AI agent will write it, deploy it, and frequently correct its own errors when prompted well. What an agent will not do is determine which problem is worth solving, supply itself with the context it needs, or recognize the failures that present as success. Those responsibilities remain yours, and they are what this course develops and assesses.
The semester therefore concentrates on the work surrounding the code: providing an agent with sufficient context to be useful, verifying its output against a reliable source, and selecting which of your ideas merits being built. HTML, CSS, and JavaScript are taught throughout as the vocabulary required to make informed decisions about what you are shipping.
By the end of this course, you will be able to:
- Explain the structure and behavior of code produced by an AI agent, and locate the source of a failure within it
- Write basic HTML, CSS, and JavaScript unaided, at the scale required to change AI-generated code deliberately
- Verify a claim about code or data against an authoritative source before acting on it
- Direct AI tools to generate, explain, debug, and refactor web code, supplying the context required for useful output
- Build and deploy responsive, interactive web applications with live URLs
- Use Git and GitHub for version control and public portfolio building
- Explain client-server architecture, the DOM, responsive design, and APIs
- Integrate external APIs and handle asynchronous data
- Evaluate competing product ideas and justify the selection of one over the alternatives
Program Learning Goals and Outcomes
This course addresses three of the Babson Undergraduate Learning Competencies.
- Problem-Solving. Analyze and assess ideas and data to make decisions and recommendations appropriate to situations and stakeholders.
- Communication. Develop and express ideas strategically, through written, oral, and visual formats, to a specific audience for a desired purpose.
- Entrepreneurial Leadership. Lead people to recognize and act on opportunities with consideration of context.
Each is developed through the projects, the in-class work, and the presentations, and is assessed as part of them.
Prerequisites
You need basic proficiency with a personal computer: using a web browser, navigating files, and installing software. No prior programming experience is required.
Course Materials
There is no textbook. All materials are provided through Canvas and the course website, and the course uses these references throughout:
- MDN Web Docs, the authoritative reference for HTML, CSS, and JavaScript
- W3Schools, beginner-friendly tutorials and examples
Required tools, all free:
| Tool | Purpose |
|---|---|
| Visual Studio Code | Code editor |
| Git | Version control |
| GitHub account (use your Babson email) | Version control and deployment |
| GitHub Copilot Free | Inline code completions, built into VS Code |
| Babson AI key | Claude Sonnet 5 for chat and agent mode in VS Code, paid for by the College |
| GitHub Desktop | Git GUI for committing and pushing |
| Google Chrome | Primary browser, with DevTools |
No paid tool is required for any assignment. You are encouraged to apply for the GitHub Student Developer Pack for its free domain name and other perks.
Optional, and demonstrated in class: Claude Code and OpenAI Codex, terminal-based coding agents included with paid Claude plans and with every ChatGPT plan. Use them if you have access. No assignment requires them.
Deliverables
Four mini projects. Each one lives in its own public GitHub repository with a live URL. They increase in scope across the semester, and each concludes in a showcase where you present your deployed site and answer a question about your own code. Full specifications and dates are published on Canvas as each project opens.
Every project carries the same required question, answered in your README and again at the showcase:
Pick one piece of AI output you did not accept as-is. What did it give you, what did you change, and how did you know it needed changing? Point at the commit.
If the agent produced acceptable output on the first attempt: what did you do to verify it?
Project 2 is built in plain JavaScript, without frameworks. A project of this size is simpler to build and to read without one, and frameworks are outside the scope of this course. An agent will frequently introduce one regardless. When that occurs, either remove it or explain in your README what it contributes and why this project requires it.
A final project. You propose it, build it, and deploy it.
| Milestone | Date |
|---|---|
| Launch and proposal workshop | Tuesday, 11/17 |
| Proposal studio | Thursday, 11/19 |
| Proposal due | Friday, 11/20 |
| Deployed URL due | Thursday, 12/03, 10:00 AM |
| Demos and voting | Thursday, 12/03 |
| Final submission | Friday, 12/11 |
Your proposal presents three candidate ideas and argues why you rejected two of them. You bring all three to the session on 11/19 and develop them in class before the Friday deadline. The deadline precedes Thanksgiving break so that you enter the break with an approved idea and can build during it. Demonstrations are held on 12/03, and the final submission is due 12/11, allowing eight days to act on what the demonstration reveals.
Voting at showcases. At every showcase you vote for one project other than your own, and write a line saying why you picked it. The line is the part that matters: it is what makes you look at the work instead of the thumbnail, and it is good material for that week's note.
Each showcase produces a class award. Winning one adds 1 point to your final grade, up to 3 points across the semester, and your final grade is capped at 100. Nobody loses anything by not winning.
In-class exercises. Many sessions produce something you build in class and commit before you leave, and some do not; when one is assigned, you will know the file by name. These are checked for presence and timing, and they are the material I draw on when questioning you during a presentation.
A weekly note. A short entry each week in logs/wkNN.md, answering four prompts:
- What you worked on
- What gave you trouble, and where you are still stuck
- Something new you picked up, in class or anywhere else
- Something you noticed about working with AI this week
In some weeks I will ask for something different. There are 14 teaching weeks, logs/wk01.md through logs/wk14.md, and a note is credited only if it was committed during the week it covers. Notes are checked for presence and timing; they are never graded on the quality of the writing.
All coursework is held in public GitHub repositories, which builds your portfolio from the first week.
| Repository | Purpose |
|---|---|
<your-github-username>.github.io |
Personal website |
oim3690 |
Course repo: in-class exercises, weekly notes |
<project-name> |
Each project in its own repo |
Commit and push frequently. Your commit history forms part of your submission and should demonstrate how the project developed. A single commit containing the entire project demonstrates nothing.
Your public work may be shown to future classes, unless you tell me at any point that you would rather it were not, or that you would rather it appeared without your name.
Class Participation
Most of the building in this course takes place in class, with me circulating among you, so attendance and sustained work during the session account for most of the participation grade.
Participation comprises three components:
In-class work and course setup (10 points). You attend and build steadily during the session, and you commit your work before you leave. A small number of short course tasks outside class count here as well, starting with the pre-course survey. Sufficient allowance is built into the measure that ordinary absence, illness, or an accommodated conflict carries no penalty here and requires no documentation.
The required 1:1 meeting (4 points). You earn the points on completion.
Lightning talks (6 points, 3 each). You give two talks during the semester, on days you choose, about something you built, discovered, or broke. Talks are given at the start of a session on a rolling sign-up; claim a slot on the Lightning Talks board when you are ready.
Each talk is two minutes, and then you run two minutes of discussion. Close with a question your classmates can genuinely disagree about, put it to the room, and call on people yourself. A talk that ends with the room silent has not finished.
Your second talk is presented from a web page you built and deployed, at your own URL. The page can take any shape you like. By that point in the semester an agent will get you there in minutes, which is the point.
I will call on students during class where it serves the discussion. These occasions are not counted and carry no points.
Grading and Course Deliverables
| Assignment | Individual/Group | Percent | Due Date |
|---|---|---|---|
| Checkpoints (7, equally weighted) | Individual | 50% | 9/13, 9/27, 10/11, 10/25, 11/08, 11/22, 12/11 |
| Project Walkthroughs (4) | Individual | 10% | At each project showcase |
| Final Project | Individual | 20% | Friday, 12/11 |
| Participation | Individual | 20% | Ongoing |
| Total Available Points | 100% |
| Grade | Range | Grade | Range |
|---|---|---|---|
| A | 94 to 100 | C+ | 77 to 79.99 |
| A- | 90 to 93.99 | C | 74 to 76.99 |
| B+ | 87 to 89.99 | C- | 70 to 73.99 |
| B | 84 to 86.99 | D | 60 to 69.99 |
| B- | 80 to 83.99 | F | Below 60 |
If you elect pass/fail for this course, a C (74) or higher earns a P. The election deadline for a full-session course is Tuesday, 11/3, at 4:30 PM ET.
Project walkthroughs
Each project showcase ends with a short individual turn: you open your own code and answer a question about it. The questions get harder across the semester, from "I point at part of your page; find the file and the lines that make it" on the first project to "say what this change will do, then make it" later.
When a question asks you to change something, say your prediction out loud before you touch the keyboard, then use whatever you normally would to carry it out, your agent included. This turn measures whether you hold an expectation of your own, because that expectation is the only thing that lets you catch an AI failure that looks like success.
There are four of these, one per project, 2.5% each and 10% of your grade together. The final project's walkthrough is part of the final project grade instead. I will go through the format in class before the first one.
What a checkpoint checks
There are seven checkpoints, equally weighted. Each one confirms that the work is being done and is on schedule. In broad terms they cover your in-class exercises, your weekly notes, whether the current project is deployed and reachable, and repository and commit hygiene.
The precise checklist for each checkpoint is published on Canvas before its due date, with the point value of every item. That published checklist is the rubric.
Checkpoints close on the Sunday that ends the week they cover, leaving Thursday evening and the weekend for outstanding work. The last one is the exception: it closes on Friday, 12/11, because it is tied to the final project submission. The first falls in week 2, early enough to identify a problem with your setup while it remains inexpensive to correct. The last is due on the final submission date and confirms that the semester's work is complete.
Four checkpoints are graded and posted before Tuesday, 11/3, which is both the last day to withdraw and the pass/fail election deadline, so that an accurate picture of your standing is available while those decisions are open.
The final project
A working web application with a live URL, accompanied by a README explaining what it does, how it is used, and what you learned in building it. Detailed grading criteria are published later in the semester, once the projects have taken shape.
Session Details
(Readings and assignments are detailed further in Canvas. Dates may be adjusted in response to the pace of the class.)
| Session # | Date | Topic |
|---|---|---|
| 1 | Tue 9/01 | Course intro, GitHub, and your first deployed site |
| 2 | Thu 9/03 | How the web works |
| 3 | Tue 9/08 | HTML essentials; Markdown and prompts |
| 4 | Thu 9/10 | Reading AI-generated code; working with an agent |
| 5 | Tue 9/15 | CSS essentials |
| 6 | Thu 9/17 | CSS layout |
| 7 | Tue 9/22 | Running a project: from an interview to a PRD |
| 8 | Thu 9/24 | Debugging with AI and DevTools |
| 9 | Tue 9/29 | Project 1 showcase |
| 10 | Thu 10/01 | JavaScript: the DOM and events |
| 11 | Tue 10/06 | JavaScript essentials (1 of 2) |
| 12 | Thu 10/08 | JavaScript essentials (2 of 2) |
| 13 | Thu 10/15 | Reviewing and refactoring AI code |
| 14 | Tue 10/20 | Project 2 showcase |
| 15 | Thu 10/22 | Arrays and loops |
| 16 | Tue 10/27 | Async and fetch |
| 17 | Thu 10/29 | API keys, .gitignore, and security |
| 18 | Tue 11/03 | Project 3 showcase |
| 19 | Thu 11/05 | Browser APIs and extensions |
| 20 | Tue 11/10 | Calling an LLM API |
| 21 | Thu 11/12 | Project 4 showcase |
| 22 | Tue 11/17 | Final project launch and proposal workshop |
| 23 | Thu 11/19 | Final project: developing the proposals |
| 24 | Tue 12/01 | Final project work and individual check-ins |
| 25 | Thu 12/03 | Final demos and voting |
| 26 | Tue 12/08 | Semester wrap-up, results, and where to go next |
Class does not meet on Tuesday, 10/13 (undergraduate fall break) or during Thanksgiving week, 11/23 to 11/27.
Religious Accommodations
Students are expected to review the Course Outline and notify faculty as early in the semester as possible of potential conflicts between course requirements and religious observances. In such an event, the faculty member will provide reasonable accommodations that do not unduly disadvantage the student.
Academic Accommodations
Any student who may need accommodation(s) based on the impact of a disability should contact the Department of Accessibility Services (DAS) as early in the semester as possible. Accessibility Services staff may be reached by email at accessibility@babson.edu, by phone at 781-239-4100, or in Park Manor West, 1st Floor. Accessibility Services staff will coordinate reasonable academic accommodations for eligible students.
Academic Integrity Policy
Academic integrity is essential to the mission of Babson College and vital to your success as a student. Our standards as an institution value community, original scholarship, and a collaborative environment that encourages each student's intellectual growth. Violations of academic integrity interrupt a student's ability to learn and produce individual work while also diminishing the achievements of the entire College. As students, you are responsible for reading, knowing, and understanding the College's policies regarding academic integrity. Below is a non-exhaustive list of potential violations:
- Cheating
- Fabrication
- Facilitating Academic Dishonesty
- Plagiarism
- Participation in Academically Dishonest Activities
- Unauthorized Collaboration
The Undergraduate Student Handbook outlines all the policies in detail as well as the potential outcomes for academic integrity violations.
What this means in a course built around AI. Using an agent to write your code is expected here and is not a violation. Handing in work you cannot explain is. If you are asked to open your code and cannot say what a part of it does or why it is there, that is treated the same as submitting someone else's work as your own. Collaboration with classmates is encouraged: discuss ideas, help each other debug, read each other's code. Copying another student's code without attribution or understanding is not that.
AI Policy
You are expected to use AI in this course. That is the subject of the course.
The line is responsibility. An agent can produce the code; it cannot be responsible for whether the result is correct, whether you can explain it, or whether it was the right thing to build. Those stay with you, and they are what is graded.
Every assignment states its own AI guidance, because it differs by task. Each assignment names one of these levels:
- By hand. Write it yourself, so that you can read the AI version later. Used for a small number of short exercises, always with the reason stated.
- AI writes, you check. This is the normal mode. Generate freely, then read every line, and be ready to explain any of it.
- You direct, you verify. This applies to the projects. Decompose the problem, give the agent context, check the result against something real, and record what you did not accept as-is.
If an assignment does not say, ask. When in doubt, be transparent about what you used and how.
Keep these in mind:
Validate before accepting. AI states wrong things confidently. In this course you will see it invent an API field that does not exist. Opening the real response was the only way to catch it.
Be careful with data. Do not paste anything personal or private into an AI tool; you cannot control where it goes. If you would not post it publicly, do not give it to an agent. Microsoft Copilot through your Babson login (copilot.microsoft.com) is a secured option.
Recognize the limits. AI output often looks better than it is. That is the failure mode this course trains you against.
Expectations differ by course. Faculty set their own rules, and following the wrong one can become an academic integrity issue. Ask when you are unsure.
Classroom Conduct Expectations
The Babson College Community Code of Student Conduct aims to promote a positive living and learning community for students at Babson. Integrity, civility, and respect for the community, in and out of the classroom, are recognized by Babson College as core values.
Babson's classrooms are spaces where students, faculty, and staff come together to engage in rigorous learning supported by integrity, respect, trust, fairness, and ownership. Every member of this learning community, including your fellow students, instructors, staff, and guest speakers, contributes to and deserves to participate in a productive environment where they can engage and learn free from disruption.
The Code's guidance on civil discourse asks students to listen actively, acknowledge others' perspectives, and address disagreements based on the structure and content of an argument or with alternative evidence. Disagreement and vigorous debate are a vital part of intellectual development and are always welcome. Disruptive Behavior, Harassment, and Bias, as defined in the Community Code of Student Conduct, are not.
These expectations apply to all course-related interactions: class sessions, assignments, online participation, electronic communications, group work, and course evaluations. Conduct inconsistent with these expectations, including conduct that disrupts a session or limits others' ability to benefit from it, may result in removal from the session and referral to the Office of Community Standards.
This restates existing College expectations. It does not create new policy.
Attendance
This course meets in person and attendance is expected at every session. Most of the work is completed in class, so an absence cannot be made up independently.
Babson expects instructors to provide reasonable accommodation for religious observance and for intercollegiate athletic competition. Notify me as early in the semester as possible and we will agree how the missed work is to be covered.
For any other circumstance that keeps you away, including illness, contact me in advance where possible. The Department of Student Advising & Success does not issue formal excuses for missed classes, so such cases are resolved directly between us.
As a matter of course policy, three or more absences without notice will prompt a conversation about whether the course remains workable for you this semester. Continued absence will be raised with your program office.
Missed Work and Late Submissions
Checkpoint deadlines are firm. Late submissions receive reduced credit at my discretion. If circumstances will prevent you from meeting a deadline, contact me before it passes. Extensions are considered case by case and require advance notice.
Laptop and Technology
Bring a laptop to every session. It may be Babson-issued or your own, and it should be current and in working order. Nearly every session involves hands-on building, and your work is committed during class.
Conflict Navigation Services
Conflict, especially when working in groups, is a normal and expected part of life, and ideally it is an opportunity to strengthen relationships and address concerns that would otherwise go unspoken. If you experience interpersonal conflict in this course, I encourage you to explore the College's Conflict Navigation Services as a resource.
Communication
The primary channels are Canvas, for announcements and grades, and GitHub, for code and notes. For any private matter, email me at zli@babson.edu with the course number in the subject line, for example "OIM3690: Question about CP2".
You can expect a reply within 24 hours on weekdays. Email received over the weekend is answered the following Monday.
Last updated: Friday, 9/25/2026