Work · PM
The Macro
Before building anything, I sat with the engineer doing the work. Most of each script was identical. A macro that fetched only the site-specific data made scripts 4–5× faster and removed transcription errors.
- Result
- 20–25 min → under 5 min
- Role
- Technical Programme Manager, Vodafone (via Interactive Management Solutions)
- When
- 2019–present
The problem
We had to produce configuration scripts for a specific type of network link. They were being written by hand, one at a time.
Watching before building
The easy route was to get the scripts produced faster. Instead, I sat with the engineer and watched how the work was actually being done, and saw a pattern. The scripts weren’t really bespoke. Most of each one was identical, and only a small amount changed from site to site.
So I stopped treating it as writing and treated it as a template. I built a macro that fetched only the variable data for each site and filled in everything else.
Result
- Time per script: 20–25 minutes by hand, under 5 minutes with the macro. Roughly 75–80% less time, or 4–5× faster.
- At 30 sites a month, that’s 7.5–10 hours of engineering time saved every month, or roughly 90–120 hours a year. Estimated from those figures.
- Transcription errors eliminated. The only manual step left was a single validation. In a live network that matters more than the time saved: a faulty configuration script puts service at risk.
- It stayed in use until the end of the programme, well after it was built.
Why it’s on this site
To be plain: this is automation, not AI. But the method is the one that comes before any good automation or AI product: watch the real work, find the repeating structure, separate what varies from what doesn’t, and automate the part that doesn’t. I got there by watching someone do the job, not by reading a requirements document.
Don’t ask people to be reliable at repetitive work. Change the system so the error can’t happen.