Cowork Lab

Cowork Lab — Following a Day at Aurora Dynamics

A seven-step lab using sample data from the fictional company Aurora Dynamics. Copy short prompts with one button and paste them right away.

Copilot Cowork Hands-on Sample data Lab guide

▶ One-line summary — Seed a tenant with 30 emails, calendar items, Teams messages, and documents from the fictional company Aurora Dynamics, then follow Ava Nakamura’s day as a sales rep and delegate work to Cowork.


Why this lab is different

The data is intentionally scattered. To find the answer, Cowork must read email, Teams, and SharePoint separately and connect them on its own.

Seeded fact Where it exists Step where it appears
The true cause of the incident One line in a Teams war-room conversation Step 4
Two toxic contract clauses RFP pages 13 and 18 Step 3
Conflict between promised SLA and actual performance Two documents in different folders Step 3
Double-booked calendar Two calendar events Step 5

You cannot find these by skimming only the beginning. The moment Cowork surfaces them is the highlight.


Fictional world setup

Everything is fictional — All companies, people, events, and numbers are invented. External domains use the reserved TLD (.example).

Aurora Dynamics — A Singapore-based industrial IoT predictive maintenance SaaS company (1,240 employees). Flagship product: Aurora Sentinel.

Role Person Purpose
Main character Ava Nakamura — Enterprise AE Login account. Owner of all data
Manager Lena Hoffmann — VP Sales Executive reporting
Technical Marcus Bello — SE RFP technical section
Delivery Priya Raman — Delivery Manager Project risk
Support Tom Okafor — Incident Commander Incident response
Legal Noah Lindqvist — Legal Counsel Contract review
New hire Alex Chen — Starts next week Onboarding

Customers — Meridian Bank (largest account) · Halcyon Energy (RFP) · Kestrel Logistics (renewal negotiation) · Solstice Retail (renewal risk)

In progress — Project Northstar (USD 4.2M, 1,870 of 3,200 devices) · Halcyon RFP (EUR 2.6M, D-18) · INC-4471 incident (six days ago)


Lab structure — Ava’s day

Each step can be run independently, but in sequence, earlier outputs become later inputs.

Step Timing What you do Duration
Step 0 — Setup Before the lab Allocate credits, deploy samples Admin 20 min
Step 1 — Morning 08:30 Email triage, priorities, unanswered replies 6 min
Step 2 — Before the meeting 13:30 Briefing from multiple sources 5 min
Step 3 — Afternoon Document work RFP analysis → deck → contract review 12 min
Step 4 — Incident response Follow-up Cross-check 3 sources, post-incident report 10 min
Step 5 — Weekly wrap-up Friday Calendar cleanup, onboarding 8 min
Step 6 — Automation Wrap-up Scheduled runs, triggers, governance 6 min

Total time — About 45 minutes, excluding setup. If short on time, Steps 2 · 3 · 4 still deliver the core.


Principles for fast results

These prompts are designed for speed first. Four rules keep demos from waiting minutes.

Principle Why
One thing at a time Three requests take roughly three times longer
Chat answers over files HTML, PPT, and Excel generation is slowest
Set numeric limits Put caps like “recent 20 emails” and “top 3”
Name the source Folders and channels remove the search step

Ask follow-ups when you need depth — Each step includes a “to go deeper” prompt. Get a short answer first, then expand only when needed.


How to run the lab

Each step repeats the same format.

  1. 📌 Scenario — Current situation
  2. 🎯 Expected output — What Cowork should produce
  3. 💡 Efficiency point — Where time is saved
  4. 💬 Prompt — Copy with the button → paste
  5. ✅ What to verify — Success criteria

Use the copy button — Click Copy prompt below each prompt and paste into Cowork with Ctrl+V.


Before you start

  • Do you have a Microsoft 365 Copilot license?
  • Has an admin enabled Anthropic models?
  • Is the tenant in the Microsoft 365 Frontier preview?
  • Have PAYG and credits been assigned? → Step 0
  • Has the sample data been deployed to SharePoint and OneDrive? → Step 0

When ready, go to Step 0 — Prepare the lab environment.


References