The numbers behind the work.
I run my personal brand with AI agents and publish the operating numbers.
What gets done. Who sees it. How much time it takes. What fails. Whether it brings business.
Latest report: Week 1
| Area | Recorded results |
|---|---|
| Operations | 3 LinkedIn posts, 5 X posts, 2 newsletter editions published |
| Distribution | 6 X followers, 6 to 14 views per X post, 249 newsletter subscribers |
| Time | About 6 hours over 3 working days, around 2 hours a day |
| Tool costs | USD 126 a month (Claude Max 100, ChatGPT Plus 20, VPS 6) |
| Business | 0 calls booked, 0 paid jobs |
| Quality | 2 agent failures recorded; minimum post score 11/18 |
What I measure
Operations
Published and scheduled work, plus the time spent running the system.
Distribution
Views, followers and newsletter subscribers.
Business
Confirmed calls booked and paid jobs.
Quality
Post review scores, failed agent actions and completed corrections.
Weekly archive
| Report | Period | Status |
|---|---|---|
| Week 1 operating report | 6 to 12 November 2026 | In progress, updated 10 November |
Each week gets its own page. Earlier reports stay available.
How I count
- Published
- Content that went live.
- Scheduled
- Content queued for publication.
- Calls booked
- Slots confirmed in the calendar.
- Paid jobs
- Jobs for which an invoice has been paid.
- Time
- My working time spent running the system.
- Agent failures
- Recorded cases where an expected agent action did not complete.
Views and follower counts are dated snapshots.
See the work behind the numbers
I document the workflows, decisions and mistakes in Experiments.