Building Real Systems with AI¶
Gonzaga University · School of Business Administration · Systems Analysis and Design
A five-week, hands-on module · November 10 to December 10, 2026
| Instructor | Tim Olsen |
| [instructor email] | |
| Office Hours | [days and times], and by appointment |
| Office Location | [building and room] |
| LMS | Canvas |
| Class Time | Tuesdays and Thursdays, [time] |
| Class Location | [building and room]. Tue, Nov 24 is online (pre-recorded videos). |
What this module is¶
In Systems Analysis and Design you learn how organizations decide what an information system should do and how it should be built: the systems development life cycle (SDLC), requirements, process and data models, specifications, testing, and deployment. Most of the time you study those steps on paper, for a system someone else will build.
For these five weeks you will build one. Using an AI coding agent as your main tool, you will take a real problem from a real user, analyze it, design a solution, and ship a working, deployed web application with a database and an AI feature. Every week is hands-on, and every week adds one of the practices that large engineering organizations use to build large systems.
You do not need to arrive knowing how to code. You will ship something on the first day that you do not fully understand, and spend the rest of the module learning what you shipped.
The five weeks¶
| Week | Theme | SDLC phases | What you build |
|---|---|---|---|
| 1 | Building with AI agents | Planning | Two working apps, one deployed to the internet |
| 2 | Analysis and design | Analysis, design | Requirements from real interviews, a spec, and a first version built from it |
| 3 | Architecture (online, Thanksgiving week) | Physical design | An architecture diagram of a real app, traced request by request |
| 4 | Data and delivery | Implementation, testing, deployment | A database with access rules, plus a team workflow with pull requests and automated tests |
| 5 | AI systems and demo day | Operation and evolution | An AI feature with an evaluation, and a live demo of your finished system |
Full dates are on the Schedule.
Learning outcomes¶
By the end of this module you should be able to:
- Explain how each phase of the SDLC is carried out when an AI coding agent writes most of the code, and what the analyst still has to do.
- Elicit and document requirements from real users, and turn them into user stories with acceptance criteria and a specification an agent can build from.
- Describe the architecture of a modern web application: where code runs, where data lives, and how a request travels between them.
- Design a relational data model, implement it with access rules enforced in the database, and integrate an external API.
- Use version control, pull requests, automated tests, and staged deployments to change a system safely.
- Add an AI model to a system as a component, and evaluate whether it works.
- Present a working system to stakeholders with evidence of whether it delivers value.
Required materials¶
A laptop (Mac or Windows) with at least 8 GB of RAM. Chromebooks and tablets will not work.
AI tools. We use Claude Code throughout the module, which requires a paid Claude account. Claude Pro at $20 per month is enough for this module, and you can cancel when it ends. If cost is a barrier, talk to the instructor in the first week; GitHub Copilot is free for verified students and works with the same project files, and Claude Code and Other Coding Agents maps every command you will learn to it.
Free accounts: GitHub, Supabase, and Vercel. Setup steps are in Computer Setup. Complete them before the first class.
All readings are on this site at no cost.
How class works¶
Class meets twice a week. Each session has a short segment of instruction and then a hands-on lab that you start in class and finish afterward. Bring your laptop, charged, to every session.
The week of Thanksgiving there is one session, on Tuesday, November 24, delivered online as pre-recorded videos. Its lab is self-paced and due the following Tuesday.
Across the five weeks you also build one project: a small information system for a real user. It runs as two sprints. The labs teach a skill; the project is where you apply it to your own system.
Grading¶
| Component | Points | % of Grade |
|---|---|---|
| Labs 1 to 8 (8 × 25) | 200 | 40% |
| Project Sprint 1: analysis, spec, and first version | 75 | 15% |
| Project Sprint 2: the finished system and final demo | 125 | 25% |
| Peer Code Review | 50 | 10% |
| Peer Product Testing | 50 | 10% |
| Total | 500 | 100% |
Every assignment is described in Assessments. If this module is part of a larger course, the instructor will tell you how these points fold into your course grade.
Late work¶
Labs may be turned in up to three days late at 10% off per day. Sprint 2 and the peer assignments fall in the last week of the term and cannot be turned in late. If something outside your control gets in the way, contact the instructor before the deadline.
Using AI in this module¶
AI is the main tool of this module, and you are expected to use it for nearly everything you build. Two rules apply.
- You are responsible for what you submit. If the agent wrote it, you still need to be able to explain what it does, why it is built that way, and how you checked that it works.
- Reflections, interview notes, and retro answers are in your own words. These are where your thinking is assessed, so do not have an AI write them.
Classroom culture¶
Expect to be stuck often. Getting unstuck is the skill: read the error, ask the agent what it means, try one change at a time, and ask a classmate before you ask the instructor. When you help a classmate, explain rather than take over their keyboard.
Laptops are open for most of every class. Keep your screen on class work, since what you show is in your neighbors' line of sight.
Getting help¶
- Ask the agent first: paste the error and ask what it means and how to fix it.
- Check this site, especially Computer Setup and the lab you are on.
- Ask a classmate or post in the course discussion board on Canvas.
- Come to office hours or email the instructor.
University policies¶
This module follows Gonzaga University's policies on academic integrity, disability accommodations, Title IX, and student conduct. See the Canvas course page for the current university syllabus statements and the offices to contact. Students who need accommodations should contact Gonzaga's Disability Access office as early as possible.