Chat-based operations
Your service is a conversation.
Your software is not.
Someone reads a message, decides what it means, then goes and does something about it somewhere else. That is human middleware. It is how cover gets arranged and how a payment gets chased, and it holds up fine until you try to grow.
Human middleware doesn't scale.
Where it lives
Chat sits in one of three states.
None of them are under control.
Invisible
Cover gets swapped between duty managers. A supplier gets chased by text. Hours get agreed in a thread that never reaches a rota. Nobody decided any of this. It just happened.
Nobody owns it and nothing records it.
Visible, unmonitored
One person sits on the group from open to close, because it is the only channel every member of staff, contractor and customer already has. The organisation knows. It has simply accepted the cost.
Known about. Never measured.
Forced into a frame
Everyone has to use the ticketing system now. So the real conversation moved somewhere the mandate could not follow, and the work carried on exactly as before. You now pay a licence for the version nobody uses.
Formal on paper. Invisible in practice.
Most organisations have all three at once, across four or five channels. An AI receptionist fixes none of it. A front desk under water is a symptom rather than the problem. The operation behind it is not working, so everything spills forward to the one number everybody knows.
Automate that and the queue goes quiet while nothing behind it improves. You have also handed away the last human voice in the building, which is the part worth keeping.
What it costs
It costs a great deal more than time.
Left alone, this caps how large you can get and wears down the people you already have.
You cannot grow past the people doing it
Every new site, corridor or contractor needs another person reading messages. The ratio never improves, so growth costs headcount at a fixed price.
Good staff spend the day being ineffective
Chasing, repeating, re-explaining, covering for someone who never replied. It fills the week and appears on no report.
Things do not get handled
A request dies in a thread. The consequence turns up somewhere else a fortnight later, and by then it costs more than dealing with it would have.
Nobody can answer the simple questions
Who is covering Saturday. Whether that payment moved. How long a trader waited. Somebody knows, but the answer lives in their phone.
How we work
We take the channels on.
Then we make them earn.
We monitor and manage conversations at scale, work out what they actually mean, and drive the action they should have triggered.
- 01
Take the channels on
We take control of the numbers and the groups, or sit alongside them using WhatsApp coexistence and the equivalent elsewhere. Nothing has to move. Nobody has to be retrained. Your people carry on messaging the way they already do.
- 02
Monitor at scale
Every channel, continuously. Volumes, response times, who is asking for what and how often. For most organisations this is the first time the chat layer has had a number against it.
- 03
Work out what it means
A shift request, a payment dispute and a broken boiler turn up in the same thread and look nothing alike. Sorting that out is the hard part. It is also the part we have spent two years getting good at.
- 04
Drive the action, then report on it
Book it, pay it, answer it, record it, into the systems that already hold your record. A request becomes a result instead of a message, and afterwards you can see what it was worth.
Any channel. SMS, WhatsApp, iMessage, and whatever your staff use internally. We go to where the conversation already is rather than asking anyone to move.
Why us
Operations people and data people,
two years on the front line.
We have spent decades inside large organisations building operational structures and the data systems that hold them together, increasingly with AI. That is the half of this problem most people never reach. Capturing chat is worth nothing if you do not know what to do with it afterwards.
For the last two years we have also run allset.chat, handling the chat operations of solo service providers every day. Payments, enquiries, bookings, and the conversations themselves.
So we are not guessing at this from a distance. We have been operating in it commercially, at the smallest unit of the economy, and we know how it behaves when you scale it up to an estate.
Chief Executive
Operations, out of professional services. Has built and run the structures this work has to land in.
Chief Technology Officer
Data and AI, out of professional services. Two decades of enterprise data systems sit behind what we build.
Chief Marketing Officer
Worked with leading consumer brands. Knows what a good conversation with a customer is worth, and what a bad one costs.
Where we work
Same job. Very different estates.
A British company, working with partners internationally. From a single service provider to a multi-site estate, it is the same job at different scales.
Start a conversation.
Tell us where the messages land and who picks them up. That is usually enough for both of us to know whether there is something here.
hello@allset.solutions