Workshop 3: Collaboration on Projects
Workshop 3: Collaboration on Projects
Use AI for meetings, project planning, shared prompts, and team documents
Tuesday, July 14, 2026 | 1:00–2:00 PM | Bring a laptop; come with a team or project in mind
No team materials on hand? Download a synthetic demo file (all content fictional and non-sensitive): Messy meeting notes (Word) · Project description (Word) · Team style guide (Word) · Update template (Word)
What You’ll Leave With
By the end of this hour, you will be able to:
- Turn meeting notes into recaps, decisions, and action items the team can use
- Use AI to draft and pressure-test a project plan
- Set up a shared Project in ChatGPT Edu—files, instructions, and chats your team works in together
- Build a custom GPT in ChatGPT Edu that your team can reuse internally
- Start a shared prompt library so the whole group benefits from what works
- Apply data-governance and consent rules to shared and team AI use
Run of Show
| Time | Segment |
|---|---|
| 0:00–0:08 | Overview: from solo use to team use |
| 0:08–0:22 | Ethics & Governance: data, consent, and equity in team use |
| 0:22–0:35 | Activity 1: Meetings and project planning |
| 0:35–0:52 | Activity 2: Create and share a Project |
| 0:52–0:58 | Activity 3: Custom GPT + shared prompt library (quick / optional) |
| 0:58–1:00 | Wrap-up and a team commitment |
Overview (8 min)
The first two workshops were about you and a tool. This one is about your team. AI helps most on shared work when it does three things: captures what happened (meeting recaps, decisions), structures what’s next (plans, task lists), and standardizes how the group works (shared prompts and custom assistants so everyone gets the same quality, not just the early adopters).
Three ideas carry the session:
- Shared Projects (in ChatGPT Edu) are a team workspace: one place that holds related chats, uploaded files, and custom instructions. Everyone invited works in the same context, so ChatGPT’s answers draw on the group’s shared files and instructions and a teammate can pick up where you left off.
- Custom GPTs (in ChatGPT Edu) are reusable assistants you set up once with your own instructions and reference files, then share with colleagues inside the Drexel workspace—so a good prompt becomes a team tool.
- A shared prompt library is just a running document of prompts that worked. It’s the cheapest, highest-return habit a team can start.
A Project is a shared workspace you and colleagues work inside together (shared chats, files, and instructions). A custom GPT is a reusable assistant you configure once and call on for a repeated task. Use a Project when a team is collaborating on an ongoing body of work; use a custom GPT when you want a consistent tool for a specific job.
For meetings specifically, Zoom AI Companion (included with Drexel Zoom) can generate the summary and action items for you automatically—a natural starting point for the recap work below. Drexel lists Zoom AI Companion as usable with High Risk Data when a meeting is not recorded or transcribed, and Low Risk Data when it is recorded or transcribed. Same principle across tools: match the tool and use case to Drexel’s data classification. See Tools & Access.
Ethics & Governance: Data, Consent, and Equity in Team Use (14 min)
The thread continues. Workshops 1 and 2 covered individual use. The moment AI use becomes shared, new questions appear: whose data is this, who agreed to it, and who has access?
Shared means visible. When you upload a file to a shared Project or custom GPT, everyone invited can see it—and the file inherits the tool’s data classification. Before adding anything, ask two questions: Is everyone in this space cleared to see this? and Is this tool approved for this data classification? A shared workspace is convenient precisely because it pools information, which is exactly why it needs more care, not less. Institutional rules still apply: data use agreements, IRB protocols, and HR/contract confidentiality don’t pause because the work is in ChatGPT.
Consent and notification in meetings. AI meeting tools record, transcribe, and summarize what people say. Tell participants when AI Companion is on, and recognize that some conversations—personnel matters, sensitive community discussions, anything under a confidentiality expectation—should not be AI-summarized at all. Notification is a courtesy and, increasingly, an expectation.
“De-identified” is not always anonymous. When teams pool data into a shared space, remember that small cell sizes, unusual variable combinations, and geography plus demographics can re-identify people—“1 case of a rare condition in one ZIP code in adults over 85” is not anonymous. Strip or coarsen identifying detail before it goes into any tool, even an approved one.
Equity of access is part of governance. If only a few staff have access to these tools or know how to use them, you create a two-tier team—and, scaled up, a two-tier school. Sharing Projects, prompt libraries, and what you learn is not just efficiency; it’s how you keep capability from concentrating in a few early adopters. This is the equity pillar of the ASPPH framework applied locally.
Before anything goes into a shared Project or team tool: Is everyone here cleared to see it? Is the tool approved for this data? Have the people in it consented to AI use? All three, or it doesn’t go in.
Activity 1: Meetings and Project Planning
Activity 1: Meetings and Project Planning
Goal: Use AI to turn raw inputs into the artifacts a project actually runs on. Use non-sensitive content or content appropriate for the tool under Drexel’s classification.
Part A — From Notes to a Usable Recap (~8 min)
Tip: if your meeting was on Zoom, Zoom AI Companion can produce a summary and action items automatically—start from that instead of raw notes.
Paste messy meeting notes (real but non-sensitive, or use Workshop3_Demo_MeetingNotes.docx — a fictional study-team check-in — or the short sample below) into ChatGPT Edu. Use GPT-5.5 if it is available; otherwise use the current default/latest GPT model in the Drexel workspace.
Use this prompt:
Turn these notes into three sections: (1) Summary in 3 bullets, (2) Decisions made, (3) Action items as a table with owner and due date. Flag anything that’s unclear or unassigned.
Notice it surfaces the gaps—unassigned tasks, vague deadlines—that are easy to miss live.
Sample notes if you don’t have your own: “talked about the survey rollout, Maria will check the IRB amendment, need final instrument by end of month, budget tight, someone should ask about the no-cost extension, next meeting in 2 weeks, recruitment slower than expected.”
Watch for (demo notes): only one decision was actually made in that meeting — did the AI invent others? Two action items have no owner — did it flag them? And there’s a personnel note that should never reach a shared recap: that’s the shared-space test in practice.
Part B — Draft and Stress-Test a Plan (~10 min)
Describe a project in a sentence or two (or paste the two-paragraph description from Workshop3_Demo_ProjectDescription.docx) and ask:
Draft a simple project plan for this with phases, key tasks, and a rough timeline. Then list the top 5 risks and what would reduce each one.
Push back: “What am I missing?” or “What would make this slip?” Use it to find blind spots, not to write the final plan.
Watch for (demo description): the description hides five real risks — a data use agreement stuck in legal review, an installation schedule the team doesn’t control, a shared analyst, student-worker coverage, and a report clock that starts regardless. How many did the AI’s risk list catch?
Reflection
- Did the recap catch anything you’d have forgotten?
- Where was the AI’s plan generic, and where did it add something useful?
Wrap-Up (2 min)
- AI helps teams most by capturing meetings, structuring plans, and standardizing how the group works.
- A shared Project gives a team one context—files, instructions, and chats—so no one re-explains the basics.
- Custom GPTs turn a good prompt into an internally shared tool.
- A shared prompt library is the easiest high-value habit to start today.
Team commitment: before you leave, name one thing your team will set up this month—a shared Project, an internal custom GPT, a prompt library, or an AI-assisted meeting recap habit.
Where to Go Next
This series is the shared foundation. Specialized sessions for research staff who code and analyze data are coming in summer or fall (AI for data cleaning, Stata/R/Python, debugging, and reproducible analysis). The series also repeats in fall.
See the Resources page for guides and further learning, and reach out to daq26@drexel.edu with ideas for future topics.