GlenndilenDigital

AI development and automation in Sussex, based in Worthing.

Practical AI for Sussex businesses: a clear first task, a pilot tested on your real examples, and a system your team can rely on, from an engineer based in Worthing.

Discuss an AI projectHow AI development works

A local partner for practical AI.

Most businesses do not need an AI strategy deck. They need one repetitive, time-consuming task handled reliably, by something that fits the tools they already use. That is where we start.

Glenndilen is based in Worthing and works with businesses across Sussex. Discovery workshops, pilot reviews and handover can happen in person, with the build, testing and support delivered remotely. You deal directly with Glenn, who designs and builds the system.

Behind the AI work sits more than ten years of software, ecommerce and integration engineering. That matters, because the model is only one part of a working system: the rest is data access, permissions, validation, integration with your existing software and a clear plan for when the AI is unsure.

Where Sussex businesses usually start.

Each starts with a clearly defined task and a way to measure success. These are the services we offer, not claims about past client results.

  • Document and order processing

    Invoices, orders, forms and emails read, checked against your rules and entered into your systems, with uncertain items sent to a person.

  • A knowledge assistant

    Answers from your own documents with the sources shown, limited to what each person is allowed to see.

  • Workflow automation

    Routine requests classified, prepared and routed through the software you already use, with approval where it matters.

  • An AI assessment

    A short engagement that ends with a written report on what is worth automating, what it needs and what it would cost. Technical consulting.

  • Private or hosted models

    Advice on whether a hosted service or a model on infrastructure you control suits your data and workload. Private AI.

  • Support after launch

    Monitoring, accuracy checks, cost tracking and improvements as your documents and processes change.

Choose a first task you can judge.

Good first tasks have a clear input, a repeatable decision and a way to check the result. Typical examples are turning supplier documents into draft product records, entering order emails into an accounts or stock system, or answering staff questions from internal guidance.

We map the current process and record a baseline first: how long it takes, how often it goes wrong and what it costs. If an ordinary rule, an integration or an existing product would do the job better, we will say so.

  • A task with clear inputs and outputs
  • A baseline recorded before anything is built
  • Conventional software where it beats AI
Workflow
  1. Request arrivesEmail, PDF or form
  2. AI extractsStructured fields
  3. Rules checkValidated or flagged
  4. Work donePosted to your systems
Illustration of an automated request workflow, not a client system.

A pilot on your real examples.

Before committing to a full build, a focused pilot tests the idea on representative examples from your own work, including awkward ones. You see where it performs well, where it struggles and what it would cost to run.

The pilot ends with a written result and a recommendation: build it, adjust the scope, or stop. Production work then adds the integrations, permissions, monitoring and support needed for everyday use.

  • Tested on your documents and questions
  • Accuracy, limits and running costs reported
  • A clear decision point before the full build
Evaluation run
Test caseExpectedResult
Clean PDF invoiceExtractPass
Photographed delivery noteExtractPass
Two orders in one emailSplitPass
Missing account numberFlagPass
Foreign currency totalConvertFix
Model AModel B
Illustration of an evaluation run, not client data.

How an engagement works.

Each stage ends with something you can review before deciding on the next.

  1. Step 1

    Assess

    A conversation and short review of the task, data and systems. You receive a written summary and a proposed pilot.

  2. Step 2

    Pilot

    A working prototype tested on representative examples, with results, limits and running costs reported.

  3. Step 3

    Build

    Integrations, permissions, review steps and monitoring, delivered in stages with your team involved.

  4. Step 4

    Handover & support

    Documentation, training and ongoing support, with accuracy and costs checked as things change.

Examples and evidence.

The interfaces and workflows on these pages are illustrations of the work, clearly labelled as such. They are not screenshots of client systems.

Client AI work is shared only with permission. If you would like to see something specific, Glenn can walk you through a working demonstration on your own kind of documents or questions, with its limits explained.

Ask for a demonstration

Questions, answered.

Anything else you would like to know? Ask Glenn directly

How do we choose the first use case?

We look for a task that happens often, has a clear input and output, and has a way to check whether the result is right. The assessment ranks candidates by value and risk so the first pilot is worth doing.

What access or data do you need?

Representative examples of the documents, emails or questions involved, and read access to the systems the result must go into. Sensitive data can be anonymised for the pilot, and access is limited to what the task needs.

How are the outputs checked?

Outputs are validated against your business rules, uncertain items are routed to a person, and accuracy is measured on a test set before launch and monitored afterwards.

Who approves actions the system takes?

You decide. Where a mistake would matter, such as creating an order or sending a reply, a person approves it. Lower-risk steps can be automated once the results have earned trust.

Which AI providers do you use?

We choose based on the task, data rules and cost, and design the system so the provider can be changed later. For sensitive work, models can run on infrastructure you control.

Who owns what is built?

You own the application, configuration and documentation built for you. Third-party models and services remain subject to their own licences and terms.

What does it cost to run?

Running costs depend on volume, the model used and hosting. They are estimated during the pilot and tracked after launch, so there are no surprises.

Do you work outside Sussex?

Yes. We are based in Worthing and work in person across Sussex where it helps, and with businesses elsewhere in the UK remotely.

Related services.

Business automation

Automate the repetitive steps between receiving a document or request and completing the work, with your team kept in control.

Explore

Knowledge systems

Knowledge assistants that find and explain the information your organisation already holds, with sources shown and permissions respected.

Explore

Technical consulting

Independent advice from people who design and build systems every day. Make the right technology decision before you commit budget to it.

Explore

Private AI

Run language models on infrastructure you control when data sensitivity, compliance or volume make a shared API the wrong fit.

Explore

Tell us about the task you want to improve.

Describe the work, the documents or systems involved and what good would look like. Glenn will suggest a sensible first step, including when AI is not the answer.

Tell us about your projectBrowse our work

Prefer email? hello@glenndilen.co.uk

Tell us what you have in mind

A few details will help Glenn get back to you with a useful answer.

Add an approximate budget (optional)

We use your details only to reply to this enquiry. Privacy. Prefer email? hello@glenndilen.co.uk