Some AI consultants advise on which tools to buy; 110 Analytics builds. We connect the systems a business already runs into one warehouse in its own cloud account, agree with the team what each number means, and put AI to work on that data: it checks every entry against those rules, flags what's missing and answers questions in plain language, showing which systems each figure came from.
What a general AI tool cannot know about your business
Ask a general AI assistant to draft a letter or summarise a long report and it does the job well, because the job needs nothing from your business. Ask it what your margin was last month and it can only work from what you paste in or upload.
Hand it this morning's sales export and it answers from that file. A colleague hands it finance's version and gets a different figure, and nothing in the tool can say which one the business agreed on. It cannot see the order still sitting in someone's inbox, the stock count kept in a spreadsheet or the customer recorded under two names.
So the hard part of AI in a business is rarely the AI. It is having one place to read from, one meaning for every number, and entries that went in properly in the first place.
What AI does on your own data
Each job reads the same warehouse your reports read, so the AI and the reports work from the same figures.
It checks what goes in
Every night, the day's entries are tested against the definitions your team agreed: a supplier invoice that differs from its purchase order, an order with no customer code, a stock level that went negative. Problems are flagged the next morning, before they reach a report.
It flags what needs you
Each problem goes to the person who owns it, with the next step beside it: a supplier's bank details changed before a payment run, a margin under plan for three months, an entry left half filled in.
It answers from your own figures
Ask the way you would ask your accountant. The answer reads your warehouse, not the internet, and shows which systems each figure came from, so it can be checked.
It drafts the routine, for a person to approve
Payment reminders, reorder proposals and replies to reviews, drafted on a trigger and queued for a person to approve. This is where AI meets business automation.
Why we connect your data before we switch AI on
We start by connecting what you already have, so you can see the whole business clearly. Then we put AI to work where it helps most.
The systems you run stay as they are. We read them; we do not ask you to change them. Each one loads every night into one warehouse in your own cloud account, and the people who use the numbers agree once what each one means: when a sale counts, what an active customer is, which site reports to which region.
The AI layer that answers questions is the last thing we switch on, not the first. It is only as good as the warehouse it reads from, so it goes on top of figures your team already trusts, and the 110 Command Centre opens each morning on the same ones.
Private AI: where your data sits
When AI answers questions about your business, it matters where the data it reads is kept, and who can open it.
The warehouse is built in your own cloud account. Each system is read under access your administrator controls, and the systems you run are never written to.
Access to the warehouse and the Command Centre is set by role before the AI layer is switched on: leadership sees the whole company, and each team sees its own part.
A general AI assistant, and AI on your own data
Both have their place. The difference is what each one reads.
| A general AI assistant | AI on your own data | |
|---|---|---|
| What it reads | What you type, paste or upload | Every system you run, loaded each night |
| Which figure is right | Whichever file it was given | The definition your team agreed, once |
| Showing its working | Depends on the tool | The systems behind every figure |
| Best at | Drafting, summarising and general questions | Questions about your own business |
A hospitality group's month, before and after
A Maltese group runs two hotels and a restaurant. Each site has its own booking or till system, the accounts sit in one system and the staff rotas in spreadsheets. The owners want to use AI, and have tried a general assistant on a few exports.
- Each site reports revenue its own way, and the assistant agrees with whichever export it was given.
- A till day with no takings recorded is found when the month is closed, if at all.
- Staff costs sit in the rotas, so cost against revenue by site is worked out once a quarter.
- A question such as which site's margin fell this month takes a day of exports and a spreadsheet.
- Bookings, tills, the accounts and the rota spreadsheets load every night into one warehouse in the group's own cloud account, with revenue defined once for every site.
- Each morning, a till day with no takings recorded is flagged to that site's manager, before it reaches a report.
- Staff cost against revenue shows by site every morning, from the same figures for all three.
- The owners ask which site's margin fell this month and get the answer in plain language, with the systems behind each figure shown.
Nobody changed the systems they work in. The difference is that the AI now reads the same agreed figures as everyone else.
From first call to AI at work
- Strategy callThirty minutes with Glen, free of charge. Bring the question you would most like answered and a list of the systems you run, and you will hear plainly what to connect first.
- Data auditAbout three hours with Glen and the people who hold and use the numbers. Every source and every report assembled in spreadsheets is mapped, and a quote closes the audit.
- BuildAccess to each agreed system, read-only wherever possible, loads every night, and your team signs off each definition. The first live screen is due three to four weeks after you accept the quote. The checks run as the figures arrive; the AI layer that answers questions is switched on once the figures underneath are agreed.
- Fractional data teamFrom the second month a senior team runs the loads and the checks, adds a new system when it arrives, and takes on the next question.
What we do not do
- We do not sell, implement or support the systems you run; your suppliers keep their roles.
- We do not switch on AI answers before the figures underneath are connected and agreed.
- AI does not make your decisions. It checks, flags, answers and drafts; your people decide and send.
- Your systems are never written to, and the warehouse sits in your own cloud account.
Glen Sultana founded 110 Analytics after a career built where the numbers move daily. Our people have run analytics for tier-one companies, S&P 500 firms and high-growth technology businesses. Glen is the person on the first call, and he runs every data audit himself.
- Built in your environment, under your account
- Every connection approved by you first
- Based in Malta
Straight answers
What does an AI consultancy do for a business in Malta?
It depends on the firm: some advise on which tools to buy, others build. We build: we connect the systems you run, agree what each number means and put AI to work on that data. Whoever you talk to, ask who does the building, and in whose account the work ends up.
Do we need new systems before we can use AI?
No. We connect the systems you already run and leave them where they are. We read them; we do not ask you to change them.
Can AI analyse our company's numbers?
Yes, if it reads the right data. On one agreed source it answers plain-language questions and shows where each figure came from; on loose exports, it answers from whichever file it was handed.
Does the AI make decisions for us?
No. It checks, flags, answers and drafts. Anything that goes out, such as a payment reminder, is queued for a person to approve, and decisions stay with the people accountable for them.
What are AI agents, and can a Maltese business use them?
An AI agent acts as well as answers: when an invoice falls overdue, it drafts a reminder in the right tone for that customer and holds it for the credit controller. It does not write to your systems, and a person sends. Agents need the same foundation as everything else on this page, connected systems and agreed figures.
What does it cost?
What it costs depends on what you need, and that is one of the things the strategy call settles. Our price follows the data audit, as a quote.
Terms on this page
- Data warehouse
- One place, in your own cloud account, where every system's figures load each night and sit side by side.
- AI layer
- The part that answers questions in plain language. It reads the warehouse, not the internet, and shows the source behind every answer.
- AI agent
- AI that acts as well as answers, such as drafting a payment reminder when an invoice falls overdue. Here, a person approves before anything is sent.
- Private AI
- AI set up to answer from your own data, read from a warehouse in your own cloud account, rather than from whatever is pasted into a public tool.