Undergraduate Session · Fall 2026 · Waltham, MA

On Oct 20 you open one system and run it.

Five Tuesday evenings. Next to your degree. Not instead of it. Not a paper. Not a chat you cannot restart on Wednesday. A named project on your laptop, standing orders in a file, and one workflow from work you already do.

  • 27 seats
  • 15 hours
  • Claude Desktop
  • $1,997 only if admitted
Example output from a live AI system built in Waltham
The class of file a student holds up on Oct 20 — from their studies or campus life, not a toy dataset.
  1. Night 1

    Sep 22

    Name the job. Create the Claude Project. Put real files on disk.

  2. Night 2

    Sep 29

    First draft lands in output/. You can point at it.

  3. Night 3

    Oct 6

    Connect one real source — Gmail, Calendar, Sheets, or search. Sandbox unless you and a faculty mentor choose otherwise.

  4. Night 4

    Oct 13

    Break it. Lock one rule in CLAUDE.md so it still works after you quit.

  5. Night five · present

    Oct 20

    Cold start. You press go. You walk the room through the file that appeared.

Sep 22–Oct 20 · 6:00–9:00 p.m. · LaCava Center · screenshot this for a roommate.

The outcome. Same for every seat.

Four things on your laptop. That is the project.

Track only names whose mess it serves — a course, an internship, a club, a family shop. The thing you present is always these four. If any one is missing on Oct 20, the step is still open.

01

The page — that is the product

A file in output/ a stranger can hold. Course Brief, Desk Panel, Weekly Run, or First Reply. If they have to open the chat to understand it, it is not done.

02

You press go

On Oct 20 you drop this week’s files and run one prompt. The page appears. The instructor points. He does not type on your laptop.

03

A named project on your laptop

Claude Desktop. Name: [Job] — [Course or org]. Folder: Documents / AI-Rocket / [Name]-[Job]. This is how you find it next week.

04

Standing orders in a file

A file named CLAUDE.md. What it does. What it must never do. Quit Claude. Open it again. It still knows the job.

Sample output · research track

This is the file. Not the chat.

SAMPLE. An HTML page the system writes to output/. A professor opens it in a browser. No PDF. No chat transcript.

output/Course-Brief-GB310.html

Open

Most of the tool is sitting unused

You own a personal jet. You're using it to go to the corner store.

Everyone in your class uses AI. Nobody can show you what it built for them. In five weeks, you will.

Why this year is different

The tool 6×’d. The homework did not.

Claude went from 60 million monthly users in January to 260 million in September. At this pace it is about 360 million by January 2027. Classmates still use it to polish a paragraph. Employers will ask what you can run — not what you can paste.

Family: this is why five Tuesdays sit next to the degree. Friends: this is the file you can send them. A future employer: this is the project that still works after the chat is closed.

If they stood on a map of the United States

  1. 60M

    Jan 2026

    East Coast

  2. 260M

    Sep 2026

    Through the Midwest

  3. 360M

    estimate · Jan 2027

    The whole country

A size picture, not where users live. 60 million is about the East Coast. 260 million reaches the Midwest. 360 million is the whole United States. Jan 2027 is an estimate on the current pace, not a promise.

What that looks like in three minutes

From something you already have. Deliver a file that speaks for itself.

Start with a real pile you already own — your notes, receipts, emails, a spreadsheet, a folder of PDFs. Build the system that turns that pile into one finished file: a report, a schedule, a tracker, a summary.

  1. 1. No fake data. If you had to make it up, it is not your pile.
  2. 2. No job-hunt chatbot. Career services already does that. Build something that does not exist yet.
  3. 3. The file has to stand alone. Hand it to a stranger with no explanation. If they cannot understand what it is and why it is useful without you walking them through the chat, it is not done.

You are not judged on the conversation you had with the AI. You are judged on the thing it produced.

Course Brief

Research or a faculty project

Before the session
Sources in five tabs. Notes in a doc no one else can follow. The next meeting starts from scratch.
On Oct 20 you do this
You drop the syllabus and two readings. A one-page brief appears: what you read, what is due, three questions for the professor. You hand them that page.

Desk Panel

Internship or work-study

Before the session
Three calendars, two inboxes, a supervisor who emails after hours.
On Oct 20 you do this
You open this week. The panel lists what is due, what is done, and what still needs you. A supervisor can read it without opening your inbox.

Weekly Run

Campus organization

Before the session
Intake in a group chat. Roster in someone’s Notes app. The weekly update nobody sends.
On Oct 20 you do this
A new name comes in. The roster updates. The weekly note drafts. You approve it. It is ready to send.

First Reply

Family shop or community work

Before the session
Messages answered when someone remembers. Follow-ups that wait until Sunday night.
On Oct 20 you do this
A message comes in. A first response drafts. A follow-up is dated. You still make the hard call. The shop does not wait until Sunday.

Who sees it

You present. Four people in the room who matter.

Night five is Oct 20. The instructor does not present for you. Faculty and family are invited. Bring classmates. A local employer can sit in.

Your professor

They read the brief or the panel. They do not have to trust a chat transcript.

Classmates

Night five is a critique. You stand up. They see the file.

Family

Invited Oct 20. They watch you press go. That is what the five Tuesdays were for.

A future employer

A named project you can restart, not a certificate of attendance.

What the session adds

Practice hours the degree does not have to carry alone.

The floor on Oct 20

You press go. A page appears. You hand it to faculty and family. That page is the product. Not a certificate. Not credit.

The stretch

Next week you drop new files and press go again. You do not rebuild the project.

A line that sits beside the degree

AI Rocket Undergraduate Session (Fall 2026). Built and presented a working AI system for [Course Brief / Desk Panel / Weekly Run / First Reply], alongside coursework. Waltham, MA.

What this is not

Not a paper the session writes for you. Not a job-hunt chatbot. Not a campus-wide app. Not a replacement for faculty or career services.

Four tracks. Pick one on the application.

The pile is yours. The object has a name.

Each track finishes in five evenings and extends work the student already does.

Course Brief

Research or faculty project

Resume
Built a living research brief and source tracker for [course / lab].
What a parent hears
They can walk a professor through the brief, not only a paper.

Desk Panel

Internship or work-study

Resume
Shipped an internship ops panel — calendar, inbox, what is due.
What a parent hears
The workplace sees they can run a desk, not only fetch work.

Weekly Run

Campus organization

Resume
Stood up intake, roster, and weekly updates for [organization].
What a parent hears
The club still moves when they are in class.

First Reply

Family shop or community work

Resume
Directed a first-response and follow-up system for a local operation.
What a parent hears
School hours went back into the family business or a community desk.

Why this sits next to the university

The degree does the long work. This session does the Tuesday work.

A university education is the foundation — judgment, writing, numbers, a field of study, and the people who teach it. That work takes years. It should. Nothing in this room replaces a professor, a syllabus, or a diploma.

The corner store is not the problem. Stopping there is. Most students already use AI to polish a paragraph. Fewer have a place, in the evening, to direct a system that helps a research project, an internship, or a campus organization — and then show that system to faculty.

This session is those practice hours. Students stay enrolled. Faculty are guests on Oct 20. The university keeps the credit. The student keeps the system.

The five Tuesdays

Evening studio. One job each night.

Sep 22 to Oct 20, 6:00–9:00 p.m., LaCava Center. Built to fit beside a full course load — not to compete with it. Claude Desktop is installed before night one so Tuesday is studio time, not setup time.

  1. 1

    Sep 22

    Name the job. Create the Claude Project. Put real files on disk.

  2. 2

    Sep 29

    First draft lands in output/. You can point at it.

  3. 3

    Oct 6

    Connect one real source — Gmail, Calendar, Sheets, or search. Sandbox unless you and a faculty mentor choose otherwise.

  4. 4

    Oct 13

    Break it. Lock one rule in CLAUDE.md so it still works after you quit.

  5. 5

    Oct 20You present

    Cold start. You press go. You walk the room through the file that appeared.

Night five · Oct 20

The instructor does not present for you.

You open the system. You walk a professor, classmates, family, and a local employer through what it does. This is a studio critique, not a replacement for a class presentation, and not a certificate ceremony.

If you cannot picture standing there on Oct 20, do not apply. The five evenings exist to earn that night.

Who the 27 are

Enrolled undergraduates. A use case from their studies.

Yes

  • Enrolled undergraduate, 18 or older, who can be in Waltham every Tuesday, 6:00–9:00 p.m., Sep 22 through Oct 20
  • Has a laptop with Claude Desktop installed — paid plan required ($30/month)
  • Brings real files from a course, lab, internship, campus organization, or family operation
  • Treats this as studio time next to the degree — not instead of it

Not this session

  • Looking for university credit or a degree requirement — that stays with your registrar
  • Wants to Zoom in
  • Cannot make all five evenings
  • Wants a chatbot chat with no file on disk and no one to show it to
Tim Regan, founder of AI Rocket

Who runs the studio

Tim Regan, founder of AI Rocket.

A Waltham business owner who built this company's operation with AI systems. He is a practitioner in the room, not a substitute for faculty.

Tuition

$1,997

Applying is free. Studio tuition for five supervised evenings — not university tuition, and it does not buy credit. Nothing is billed until you accept the seat.

If you are a parent

This is not a Bentley course and not for credit. Your student stays enrolled. On Oct 20 you can sit in the room and watch them present a system they built on their own laptop. The application asks who pays — student, parent, split, or sponsor — so that is settled before a seat is offered.

The fine print, up front

What do I actually walk out with?

A page that appears when you press go — a Course Brief, Desk Panel, Weekly Run, or First Reply. Plus the Project and CLAUDE.md so you can run it again next week. The university keeps the credit. You keep the machine that prints the page.

Is it an app I publish?

No. It lives on Claude Desktop. You name it, you run it, you show the file it produced. If it only works while a chat is open, it is not done.

Does this replace a university course?

No. Students stay enrolled. This is an evening studio — five Tuesdays next to the semester, not a substitute for it. Credit, majors, and graduation requirements stay with the university.

Is this a Bentley program?

No. The room is in the LaCava Center at Bentley University. The session is run by AI Rocket, an independent Waltham studio. We are grateful for the campus setting. We are not a university department and we do not speak for the institution.

Will faculty be involved?

Faculty are invited to night five on Oct 20. Students are encouraged to choose a pile a professor would recognize — a research thread, an internship, an organization they already serve.

Do I need to know how to code?

No. You work on your own laptop. What you need: comfort in a browser, real files from your studies or work, and Claude Desktop installed before night one.

Where is it? Parking?

LaCava Center at Bentley University, 175 Forest Street, Waltham, MA 02452. Parking on site. Reachable from Boston by commuter rail or car.

What do I need to bring?

A laptop with Claude Desktop installed before Sep 22, and a paid Claude plan ($30 per month). The day you are admitted you get a short setup note so Tuesday evening is studio time, not install time.

Download Claude Desktop · See Claude plans

When do I pay?

Only if admitted. Applying is free. Nothing is billed until you accept the seat. Tuition is $1,997 for the five evenings.

What if I miss a Tuesday?

The same 27 people every night. All five evenings are the deal. If the semester already owns those hours, do not apply — a half-used seat is a seat another student needed.

Apply

Pick a track. Four questions. No essay.

Tuition is $1,997 — billed only if admitted. Name the project you will present on Oct 20.

  1. Track. Research, internship, campus org, or family / community shop.
  2. Sentence. What will that system do by Oct 20?
  3. Tuesdays. All five, 6:00–9:00 p.m.
  4. Payer. I do · A parent · We split it · Someone else (employer, sponsor)

How you get the seat and the building. Nothing else.

Selected students. Closed room. No fee to apply. Nothing is billed unless you accept a seat.

By submitting you agree to our Privacy Policy and Terms. We reply by email or phone about your request only.

The line that sits beside the degree

AI Rocket Undergraduate Session (Fall 2026). Built and presented a working AI system for [use case], alongside coursework. Waltham, MA.
Apply for the session

27 seats. You present Oct 20. Starts Sep 22.

Undergraduate Session · Tim Regan · AI Rocket · Waltham, MA · tim@airocket.pro

An independent studio hosted in the LaCava Center. Not a university course. Not for university credit. Offered as an extension of a student's studies.

Apply free — run it Oct 20