Case study
- Client
- A residents' forum in the south of England (placeholder)
- Sector
- Community
- Capability
- Agents
- Date
- May 2026
Agents for Change
Placeholder — illustrative sample, not a real engagement
TL;DR
This is placeholder content — a stand-in until a real anonymised write-up exists. In the sample scenario, a residents' forum representing a made-up neighbourhood of 9,000 people collected reports of local problems in chat threads nobody had time to sort through.
We're imagining a simple agent on the messaging app the group already used: members send a report the way they'd message a neighbour; the agent sorts, transcribes and groups it; a weekly summary goes to the relevant authority. Every figure here is invented; replace this file before launch.
The work
The constraint
This section is placeholder text. Imagine a fictional community group that represented a made-up neighbourhood of 9,000 people and campaigned on invented local issues.
For the purposes of this sample, the group’s evidence was scattered across messages, photos and voice notes that nobody had time to sort through.
What we built
Placeholder narrative. The fictional team is said to have built a simple agent on a messaging app the group already used, so members could send a report the same way they would message a neighbour. There were no forms and no new app to install in this imaginary version.
Picture the invented agent sorting each submission by issue type, pulling out useful details, transcribing voice notes and grouping everything so the group could see patterns. Once a week, an imaginary summary went to the relevant authority.
Who runs it now
More placeholder prose. In the sample, two volunteers on the committee own the categories and the weekly send; nothing needs an engineer.
None of this reflects real work. Delete this file and write the genuine case study when it is ready.
The impact
Placeholder framing line. In this fictional scenario casual messages became something the group could act on.
Made-up figure. The imaginary agent handled around 800 reports from members over a pretend six months, with no forms and no new app to install.
Invented outcome. That fictional evidence supposedly supported two notional campaigns with the local council.
Closing placeholder line. Swap this entire section for real impact data once a project is documented.
Got a problem that looks like this one?
Tell us where the work piles up. We'll tell you honestly whether we can move the number.