A national industrial services company with sites around the country was running its day across nine shared mailboxes and two systems that did not talk to each other. StellaIQ built one screen to run it, cutting manual document processing by roughly 80% and, in the first fortnight, finding about £15,000 of deliveries with nothing to prove they ever arrived.
Nothing dramatic. Just a day spent moving information between places.
Customer email arrived in nine shared mailboxes, and nobody could see across them. Orders lived in one system and deliveries were tracked in another.
Proof of delivery arrived as PDFs that someone had to open, read and file. The office spent its day moving information between places rather than acting on it.
None of it needed new software. It needed the systems they already had — an ERP, HubSpot and Microsoft 365 — joined up behind one layer.
One screen to run the day, built around the systems they already had.
Waiting every morning: what came in overnight, what has already been dealt with, and the handful that need a person.
All nine mailboxes come into one place, and AI sorts every email into a category. Complaints always come in as urgent.
Routine order emails are answered automatically, in under a minute, in the company's own words. Anything else goes to a person.
The proof-of-delivery paperwork that arrives by email is read and matched against what was ordered, so gaps show up straight away rather than at month end.
Manual document processing fell by roughly 80%. In the first fortnight, the delivery check found about £15,000 of deliveries with nothing to prove they ever arrived. The system went from kickoff to live in fifteen consultant-days.
Roughly 80% less manual document processing.
Of deliveries with nothing to prove they ever arrived.
Consultant-days to a working system.
Both, each where it belongs.
AI does the reading: it sorts incoming email and pulls the details out of delivery paperwork, which arrives in whatever layout the sender chose. Rules do the checking: matching a delivery to an order is arithmetic, so it is instant, auditable, and cannot invent an answer.
When the system is not sure, it asks a person instead of guessing. Everything runs in the company's own Microsoft cloud, documented, with the client's own administrator access.
If your day runs through shared mailboxes, paperwork that arrives by email, and systems that do not talk to each other, the pattern is the same. The documents change — supplier statements for a wholesaler, job sheets for an installer, CVs for a recruitment agency, matter files for a law firm — but the job is the same: read what arrives, put it where it belongs, and ask a person when it is not sure.
We do not publish client names, and we will not publish yours either.
No. It connected the systems they already had — an ERP, HubSpot and Microsoft 365 — behind one automation layer. The team kept working in the tools they knew.
Only routine order emails get an automatic reply, in the company's own words. Anything else, complaints included, goes to a person.
It runs in the company's own Microsoft cloud under their own administrator access. Only named staff can sign in, and nothing is used to train anyone else's model.
Fifteen consultant-days from kickoff to a live system.
We do not publish what individual clients pay. Engagements start with a £2,500 operations audit; a first automation build starts at £5,000, and full engagements at £12,500, all plus VAT.
Forty-five minutes, no obligation, and an honest answer about which of your problems is worth automating and which is not.
Last updated
We'd like to use analytics and advertising cookies to see how people find this site. Nothing is set unless you agree. What we collect.