Getting more from the systems you already run

Most businesses are not ready for AI, and the reason is rarely the technology. The work itself is not tidy enough yet: the steps change from person to person, the rules live in people's heads, and nobody has written down what a good result looks like. Enterprise AI pilots still fail at high rates, and the cause is usually the foundation underneath the tool rather than the tool itself.

The value rarely needs a new platform. Most of it is already in the systems you pay for: the shared inbox, the CRM, SharePoint, WhatsApp and the spreadsheets that grew up around them. AI can read, sort, answer and route the work your people currently do by hand, write the results back into those systems, and pass anything that needs judgement to a person. The skill is in finding the right place to start, and sometimes waiting is the right call.

Below are four use cases from four different businesses. The systems and the problems differ, but each followed the same three steps. First, understand the problem as it really runs day to day. Then scope the smallest build that proves the value. Then deliver it into the tools people already use.

Documents that file themselves

Understand. Invoices, claims, onboarding packs, contracts, delivery notes and tax documents arrive as password-protected PDFs, phone photos and scans bundled several to a file. In most companies a skilled, well-paid employee still opens each one, works out what it is and whose it is, and retypes the contents into the system. That time is the cost, and it peaks exactly when the business is busiest. Watching that person work is the fastest way to find the real specification, because the rules they apply in their head are the rules the system will need.

Scope. Anything that is really a business rule is decided up front rather than left to the AI to guess. For example, a document that contains another document is still the document it is. The list of document types is held as plain data the business owns, so adding a new type is an admin task rather than a rebuild. A document the system cannot place is never filed on a guess. It becomes a visible "needs review" item for a person, with the system's best guess attached. We also pick the AI model on evidence: candidates are graded against hand-checked examples, and in one test a mid-priced model matched the most expensive one on accuracy at 2.3 times lower cost.

Deliver. For one client, we built an intelligent document extraction pipeline that has now processed more than 100,000 of their documents. They arrive by email, chat, client portal and a watched folder. Each one is unlocked if it is password-protected, read, matched against the client's own list of document types, and has its main details pulled out and saved as a tidy record in the system the business already uses. We measured accuracy and cost as we went: in one run for that client, covering tens of thousands of records, accuracy checked by hand came to 99.3%, and the computing cost for the entire run was about $216.

What changes for the team. The details pulled from each document keep client records current, feed reporting and flag the cases that need a conversation. Nobody opens every attachment any more; the team works from the short list of documents the system was unsure about. Because these pipelines handle sensitive records, we design data security into the architecture from the first day.

A CRM that keeps itself up to date

Understand. Client records go stale for a boring reason: updating them is a chore. A new proof of address arrives, someone files it, and the address on the record stays the old one. Six months later nobody trusts the data, and the CRM the business paid for stops being the place people look first.

Scope. For an advisory business, we put an assistant in a panel right beside the client record. We started strict, with a handful of document types: the ones the team handles most often rather than the most complicated, because the value comes from volume. Each type sets out which details to read, which folder the file goes to and what it is called. Nothing is saved until a person confirms it, so a named human stays accountable for what is filed. Handling documents that arrive with no client attached was scoped as the next step, with its own price.

Deliver. A document is dropped into the panel. The assistant works out what it is, reads the details that type calls for, and shows them next to what the record currently holds. The person's job shrinks to one glance: anything that has changed is highlighted, and one click files the document and updates the record.

What changes for the team. Before, updating a record meant opening the document, reading it, retyping the details and filing it by hand, which is why it often did not happen. Now a person looks at one highlighted difference and clicks once. Every document is filed under the same name, tags and folder as every other document of its type, so anyone can find what a colleague filed. Where a new value disagrees with what is already held, both are shown side by side, and nothing is overwritten without someone seeing it. Because the list of document types is data, an administrator can add a new one with no development work.

Customer conversations on WhatsApp

Understand. For many businesses, WhatsApp is where customers already are. It is also where repetitive questions, after-hours messages and promising leads pile up, leaving the business to choose between hiring more people and letting enquiries wait.

Scope. An intelligent WhatsApp agent answers only from the company's own approved material and can take only the actions on an agreed list. Every conversation, document and qualified lead is recorded in the CRM the business already runs, such as Microsoft Dynamics 365. Questions that need an expert go to the right person with the full conversation attached, and the same pattern extends to Microsoft Teams and the website.

Deliver. Our launches typically begin with a soft pilot and staff on standby, so the business sees real conversations before it relies on the system. Over two months in production at a financial services firm, one of our agents handled 8,188 real cases, resolved 90.6% without a human agent and returned 1,154 hours of staff time to the team, about 577 hours a month, at roughly $0.09 per case.

What changes for the team. Customers get an answer at 11pm on a Sunday. Staff spend their day on the conversations that need them, with the history already in front of them, and every lead lands in the CRM instead of on someone's phone.

Spreadsheets and licences, in one place

Understand. In a recent Blueprint for a large UK property company, individual staff had signed up for their own AI tools to get through the day. That helped them personally but gave the business no shared standard or platform. The company was paying roughly £40,000 a year in licences for fairly basic analytical work, and senior team members spent 147 hours a month on high-volume manual tasks across spreadsheets with no single source of truth.

Scope. The build replaces the licensed tools and pulls the work into one place, with direct connections to the specialist platforms the company already relies on, so data flows without being retyped.

Deliver. The rebuild replaces the licensed tools for about £2,000 a year in hosting. Manual effort is projected to fall to 18 to 23 hours a month, human review included, which returns 124 to 129 hours every month.

What changes for the team. Senior people get most of a working week back each month, the business owns one platform instead of a dozen personal subscriptions, and everyone works from the same numbers.

What the four have in common

The systems and teams were different, but the method was the same. Each began with a process that already had clear rules, high volume and a cost the business could measure, because the return depends on choosing the right problem. Each was scoped small enough to ship quickly and big enough to matter. Each wrote its results back into a system the business already ran, and each kept a person in charge of anything uncertain.

They also passed the same readiness test. If the person doing the work can explain the rules, a machine can follow them. If they cannot, there is a process problem to fix before AI can help, and the honest answer is to fix that first.

That method is the AI Blueprint

Understand, scope, deliver is how our AI Blueprint works. It starts with a four-week Readiness Sprint that maps how your processes actually run, identifies the highest-value opportunities and picks the one worth building first. The build phase that follows puts working AI directly into the tools you already use, at a fixed price.

You leave with a working prototype of your highest-priority use case, a person approving anything uncertain, a full record of what the system did and why, and a training and rollout plan. It suits organisations that want working AI in their day-to-day tools without waiting months for separate discovery and build phases. And if the honest answer is that you are not ready yet, we will tell you that too.

The Claude Partner Network

On 2 October 2026 we were certified in Anthropic's Claude Partner Network, which extends what we can offer to company-wide Claude rollouts, new agents for WhatsApp, Teams and the web, and Claude embedded directly in systems such as Dataverse, CRMs and service desks. For our clients, each of the use cases above can now extend across more of the business.

FAQs

What does "getting more from the systems you already run" mean in practice?

It means adding AI to the tools your business already pays for, such as your inbox, CRM, SharePoint and WhatsApp, rather than buying a new platform. The AI reads, sorts and routes work inside those systems and writes the results back, so your team keeps working where they already work.

How do I know whether my business is ready for AI?

Ask the person who does the job today to explain the rules they follow. If they can, a machine can follow them too. If the answer changes from person to person, the process needs tidying first, and the honest answer is to fix that before adding AI.

What is intelligent document extraction?

Software that takes a document in whatever form it arrives, such as a scanned PDF, a phone photo or a password-protected file, works out what it is, pulls out the details that matter and files them in the right place. Anything it is unsure about is flagged for a person rather than filed on a guess.

Do we have to replace our CRM or other systems?

No. Every use case in this article wrote its results into a system the business already ran. The aim is to make the systems you have more useful.

What happens when the AI gets something wrong?

Uncertain results go to a person. A document the pipeline cannot place becomes a visible "needs review" item, a CRM update waits for someone to confirm it, and a WhatsApp question the agent cannot answer is handed to a staff member with the full conversation attached. Nothing is overwritten without someone seeing it.

How accurate is it?

We measure it on real records. In the document extraction run described above, accuracy checked by hand across tens of thousands of records came to 99.3%. Every build also keeps a human check on anything the system is not confident about.

How long does an AI Blueprint take?

The Readiness Sprint takes four weeks and ends with the use case worth building first. The build phase that follows is delivered at a fixed price into the tools you already use. Details are on the AI Blueprint page.

Where should we start?

With the process that has clear rules, high volume and a cost you can already measure. It is usually the one your most experienced people complain about most.