Proof
What's running, and how we build it.
WinSphere builds the automation that takes over the work nobody gets to, for owners of firms and agencies with 10 to 100 people: first we run the numbers, then we build, then we keep it running. WinSphere is an AI-first agency: we do not just build with AI for clients, we run on it ourselves. Below you see what of that is running today.
What's running right now
50
conversations with owners of firms and agencies about where work piles up
3
systems running today, not on a drawing board
05:00
every day again: the exclusion list that stops anyone being approached twice
What's running
Three systems, and with each one the number we actually have.
Runs daily
Our own operation. WinSphere itself runs on its own automation, from the first lead list to the invoice.
Ourselves. Whatever we build, we use first.
Shows that a system keeps working once the first week is over, not just on handover day.
Runs daily at 05:00
Our own exclusion list. The idea: anyone who has ever replied should not be approached again, not even from another channel. The result: every day, new replies from HeyReach, Instantly and HighLevel are merged into one clean list, so nobody is approached twice. The source: our own execution logs. Runs daily at 05:00.
Ourselves: our own outreach to firms and agencies.
Pulls in the new replies from three systems every day, merges them and keeps one clean list, without a manual step.
Running at the client
Sourcing agent for a recruitment agency. Vacancy in, scored candidate list in the CRM out; the manual work of hours of searching disappears.
A recruitment agency, anonymous at the client's request.
From manual sourcing to a scored shortlist in the CRM, in minutes.
How we go about it
The same order with every system above.
Not the luck of a good tool, but a fixed way of working.
01
First we run the numbers.
Before we build anything, we work out on the actual figures whether it pays off. No assumption, no pitch.
02
Then we build.
In four-week cycles, with a baseline up front, so you don't have to believe it works, you see it.
03
Then we keep it running.
We stay on the system: measuring, adjusting, standing behind the outcome.
What you can see today
Two things you can already use
12
The twelve tasks nobody gets to most often
Twelve recognizable tasks drawn from more than fifty conversations with owners. Tick what you recognize and see what one of them costs per year.
Take the check
8
The eight questions we ask in the first thirty minutes
The same eight questions that are on the open client requests page; they work for any firm or agency, not just accountancy.
See the eight questions
Who it's for
For owners who want to see what is actually running first.
Not a fit
You are looking for a demo or a standalone chatbot.
You decide on logos and references only, not on a costed plan built on your own figures.
You want to overhaul your entire process first.
You already have someone whose job it is to build these systems and keep them running.
A fit
A firm or agency with 10 to 100 people, where work nobody gets to piles up every week.
You would rather have a number worked out on your own figures than a wall of logos.
You would rather judge a baseline and a follow-up measurement than a wall of logos.
You actually want this: someone on your team gets genuinely excited about doing work differently with AI and automation.
Everything above is running today: our own operation, an exclusion list that runs every day at 05:00, and a search agent at a search firm.
WinSphereAI and automation for firms and agencies with 10 to 100 people. First costed out, then built, then kept running.