AI for UK SMEs.
Identified, planned, and implemented.

AI isn't coming. It's already in the tools you use, and the gap between businesses that adapt and businesses that don't is widening fast. I find the AI work your tools can't reach, plan how to build it, then build it for you.

Drawing on 15 years across UK financial services, regulated project leadership, and NuFlw, my AI product running live on the Claude API.

15+ years inside UK SMEsPRINCE2®-certifiedNuFlw, an AI-powered SaaS running on the Claude API

Where does AI fit in your business?

Answer five questions. Get a personalised preview of the opportunities a full AI Process Audit would surface for a business like yours.

Step 1 of 520%

Step 1 of 5

About your business

A couple of quick basics so the preview lands in the right ballpark.

Industry

Team size

Three things I work on. All in public.

01

See where AI saves you time

I find the parts of your business losing hours to manual work, design the fix, and build it. You stop reading about AI and start watching it pay back.

See the methodology

02

NuFlw

My marketing-operations SaaS, built on the Claude API and in use today. Proof that I don't just advise on AI, I've built it. The same approach I bring to client work.

Visit NuFlw

03

What I've learned, in writing

Notes on running operations, fixing broken processes, and getting AI working in real businesses. Built from 15 years of doing it.

Read the blog

There is no shortage of AI consultants.

Most of them learned AI in 2024. A ChatGPT subscription, a Zapier account, and confidence in a discovery call. That is the market.

Here is what is different about working with me:

I spent seven years as a regulated UK mortgage adviser. I spent three years as Project Manager at Bath Building Society, mapping end-to-end customer processes across a regulated institution, building digital infrastructure, managing offshore developers, and reporting to the board.

Then I built NuFlw: an AI-powered SaaS running on the Claude API, in production, used by real clients today.

I do not connect your tools with a workflow builder and leave. I map your operations, identify where AI creates genuine leverage, build the system, train your team, and stay in as fractional AI lead until it is embedded.

The difference is not a slide deck. It is documented, measurable, operational change.

How AI works in an operationally complex business.

Not a platform. Not a tool. A layer of intelligence built on top of how your business already operates.

EmailDocumentsWebVoiceExisting Systems
AI Operations Layer
Built on your existing workflows. No rip-and-replace.
Document Processing
Extract, classify, and route incoming documents automatically.
Productivity
Client Communications
Draft and review client-facing messages within your compliance guardrails.
Service
Compliance Checking
Flag issues before they reach a human. Audit-ready records by default.
Compliance
Reporting and Analytics
Structured summaries and performance data without manual compilation.
Analytics
Workflow Automation
Trigger actions across your existing tools based on defined business rules.
Operations
Hours recovered each week
AI removes the manual work from your team's week.
Fewer manual errors
Structured outputs mean fewer human errors.
Faster client turnaround
Automated workflows reduce delays at every stage.
Audit-ready records by default
Every AI action is logged and reviewable.
Client Work

Shields Buildings: AI-powered lead scoring for a bespoke garden buildings business.

Shields Buildings designs and builds bespoke garden rooms and outbuildings across the South West. The business runs on a CRM and had a large, unstructured contact database with no consistent method of identifying which enquiries were most likely to convert.

What was built

A lead scoring engine that analyses CRM contact behaviour, assigns weighted scores based on conversion signals, and produces a daily digest email ranking warm leads for follow-up. The engine runs automatically. The sales team starts each day with a clear, prioritised list rather than a full database to work through.

A weekly warm-up list runs alongside it: 24 long-quiet contacts who have recently re-engaged, each with a personalised hook, suggested subject line, and opening line. Built to be worked in a single focused session.

"We went from a full database to a prioritised list overnight."

Paul Gent, Shields Buildings

Shields Daily · Tuesday 16 June
5 priority leads · 71 active engagements · 24 still-engaged warm-ups
107
Neil Rogers
Quoted, visited display, downloaded brochure. Phone call today.
75
Kevin Richards
Quiet for 417 days, now clicking again. Send personalised email.
73
Barrie Bailey
Quoted, 4 opens in 30 days. Direct nudge on quote overdue.
70
Gilbert Hayward
Quiet for 345 days, 2 clicks in last 7 days. Re-engagement pattern.
69
Matt Sexton
Quoted, closing email triggered. Phone call today.
Generated automatically · domleigh.com
Pipeline snapshot
11
Currently quoted
3
Site or display visit
20
Quote enquiry submitted
832
Brochure in hand
Built by me

NuFlw: what I build looks like this.

NuFlw is a marketing-operations SaaS I built on the Claude API. It is live, in production, and runs real marketing operations for real clients today. This is not a demo. This is what AI implementation looks like when it is done properly.

NuFlw Forge Intelligence dashboard
Click to expand
The Interface

A full marketing-operations layer.

NuFlw gives marketing teams a single interface for managing AI-assisted campaigns, lead pipelines, and content operations. Built on the Claude API with structured output parsing throughout.

The Agent Layer

AI agents doing the operational work.

NuFlw uses AI agents to process marketing data, classify leads, and trigger follow-up sequences. Every agent action produces a structured, reviewable output. Nothing black-box. Every decision is logged.

NuFlw projects workspace
Click to expand
NuFlw client onboarding pipeline
Click to expand
What This Means For Client Work

Proof, not a pitch.

When I tell a client I can build an AI layer into their operations, NuFlw is what that looks like in practice: defined inputs, structured processing, reviewable outputs, embedded into existing workflow. The same approach, applied to your business.

See NuFlw
Dom Leigh

Three years as Project Manager at a National Building Society delivering digital transformation in a regulated environment: mobile app, Broker Origination System, API integrations, and a Company Process Improvement project that mapped end-to-end customer journeys to identify and remove operational bottlenecks.

Before that, seven years as a regulated mortgage adviser, understanding how UK financial services SMEs actually run, from the inside.

Since 2024, building marketing and operations systems for UK SMEs that have delivered measurable outcomes: a 125% lift in inbound leads for one client, full brand and ecommerce builds for others.

Today, founder of NuFlw, an AI-powered marketing-operations SaaS running on the Claude API and in use today. PRINCE2®-certified throughout. The same process-mapping discipline that identifies bottlenecks in a regulated building society is what identifies AI opportunities in your business.

See the diagnostic methodology in action.

An interactive walkthrough of the process mapping approach behind every AI Process Audit. Nine stages, six bottlenecks, and a clear view of where AI fits in any operationally complex business.

Explore the case study
Intake
Verification
Review
Handover

Where does AI fit in your business?

Start with an AI Process Audit. 5 days, fixed price, written report and a 90-minute walkthrough. Not a sales pitch. A diagnostic.

Book your audit, £495