Codex to fit business travel into a calendar
Dr. Eva-Maria Hempe tested Codex against an actual travel requirement: reaching Frankfurt by 10:30 on June 29. A comparison of flights, trains and driving needed to include the journeys to stations or airports and the final leg to the meeting location.
linkedin.com
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Compare car, rail and flights for my Frankfurt meeting arrival by 10:30 on June 29. Include first and last mile, propose calendar blockers, and ask before moving any confirmed meeting.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Dr. Eva-Maria Hempe
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- The test produced route options and preferred-route calendar blockers, plus approval-based meeting rescheduling.
- Human involvement
- Hempe approved meeting moves and selected what to keep or cancel; no autonomous ticket purchase is described.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- State the required arrival time and use the calendar's meeting location.
- Compare road, rail and air options including first and last mile.
- Block the preferred route and approve decisions about affected meetings.
Dr. Eva-Maria Hempe tested Codex against an actual travel requirement: reaching Frankfurt by 10:30 on June 29. A comparison of flights, trains and driving needed to include the journeys to stations or airports and the final leg to the meeting location.
The meeting location came from her calendar, while reusable personal context described details such as traveling with hand luggage. Rather than repeatedly entering that information, she could use it across planning tasks. An unnamed workaround let the preferred travel option become calendar blockers, though the post does not document ticket purchase.
Her calendar experiment also covered an upcoming vacation. Codex cleared tentative meetings on the specified days, then presented other meetings for a decision. Hempe could keep, cancel or move them. When moving a meeting, the workflow looked for times suitable for the attendees and sent an update only after her approval.
The combination makes this more specific than a generic itinerary: route feasibility and calendar consequences were handled together. Her airport timing preference was personal context for the test, not a universal recommendation. She also notes integration and permission limitations. The reported output is a travel comparison and blocked calendar time, with explicit human decisions on the remaining meetings; the account does not establish a booked or completed Frankfurt journey.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: www.linkedin.com · Dr. Eva-Maria Hempe
Frequently asked questions
Were station and airport transfers included?
Yes. The actual request included first and last mile and the meeting destination from the calendar.
Could it send a rescheduled meeting without approval?
The described workflow asked for Hempe's approval before sending the updated invitation.
Was a ticket booked?
The account describes planning and calendar blockers, not a completed ticket purchase or journey.
More details
About this account
Codex compared transport options and managed calendar planning with reusable personal context.
Firsthand source opened and read. Suggested prompt written by us from the reported workflow; no original task prompt is quoted.
Credit: Dr. Eva-Maria Hempe
Source published
Added to Agents In Use
Source checked
Tools and topics
Codex · Calendar · Persistent personal context
Business travel · Calendar blockers · First and last mile
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.