Frequently asked questions

An AI agent is software that does a task for you, from start to finish. It reads information, makes simple decisions and takes action. For example, it can read a customer email, find the order, write a reply and send it. You set the rules. The agent follows them, day and night.

We want you to see the value before you spend any money. Most business owners do not know where AI can help them. The blueprint shows you, with real numbers. Many clients then choose us to build the system. If you do not, you still keep the blueprint.

No. There is no contract and no obligation. You can build the system with us, build it with someone else, or do nothing. We will not chase you with hard sales calls.

ROI means return on investment: what you get back compared to what you spend. We look at each task and measure the hours your team spends on it, the cost of those hours, and the cost of errors or delays. Then we compare that to the cost to build and run the AI agent. We show every number and every assumption, so you can check them yourself.

We follow UK GDPR, the UK law that protects personal data. Your data stays in UK data centres. We give each AI agent access only to the data it needs for its task. We sign a data processing agreement with you, which sets out how we handle your data.

We design every system with checks. For important actions, such as sending a quote or a payment, a person on your team approves first. Every action is logged, so you can see what the agent did and why. If we find an error, we fix the system so the error does not happen again.

That is your choice. Most of our clients use AI to remove repetitive work, not people. Your team then spends more time on customers, sales and the work that needs a human. As your business grows, you can take on more work without hiring as fast.

A chatbot answers questions. An AI agent does work. It can log into your tools, update records, send emails and create documents. A chatbot can be one part of an agent, but an agent goes further.

Normal automation follows fixed rules: "if this happens, do that". It breaks when the input is messy, such as a free-text email or a scanned PDF. AI agents can read and understand messy input. We often combine both: simple automation for fixed steps, and AI for the steps that need judgement.

AI agents work best on tasks that repeat often and follow a clear pattern. Common examples are: answering customer emails and enquiries; following up with new leads; booking appointments and sending reminders; reading invoices, receipts and forms, and entering the data; creating quotes, proposals and reports; updating your CRM (customer database) after calls and meetings; and sorting and routing support tickets.

Tasks that happen rarely, change every time, or need deep human trust are a poor fit. Examples are complex negotiations, sensitive staff matters, and one-off creative projects. We will tell you honestly if a task is not worth automating.

We work with small and medium businesses (SMEs). Our clients include professional services, e-commerce, property, trades, healthcare clinics and logistics. If your team does the same tasks again and again, we can probably help.

Probably not. If one person spends more than 5 hours a week on repetitive admin, automation can pay for itself. The free blueprint shows you if it makes sense for your business. If it does not, we will tell you.

No. We handle all the technical work. You only need to explain how your business works today. We train your team to use the new system, in plain language.

Usually only a little. We build the agents around your current process and tools. Sometimes we suggest a small change that makes the process simpler. You decide if you want it.

No, and you should not want that. We focus on the tasks where automation gives the biggest return. Your people stay in control of decisions, relationships and quality.

We work in 3 steps. 1. Discovery (free): we talk with you and your team to understand how your business runs and where time is lost. 2. Blueprint and ROI (free): we give you a written plan of every task AI can automate, with the expected savings for each one. 3. Build (optional): if you want, we build, test and launch the AI agents for you.

The first call takes about 30 to 45 minutes, on video. We ask about your business, your team, your tools and your biggest time-wasters. You do not need to prepare slides or documents. At the end, we agree if a blueprint makes sense for you.

In total, about 2 to 3 hours of your time over 1 to 2 weeks. That is usually the first call, one deeper session to map your processes, and one meeting to present the blueprint. Sometimes we also speak for 15 minutes with team members who do the work every day.

Nothing is required. It helps if you think about 2 or 3 tasks that take too much time, and roughly how many hours they take each week. A list of the software you use is also useful.

The blueprint is a written document. It contains: a map of your key business processes; a list of tasks that AI can automate, ranked by value; how each AI agent would work, in plain language; the expected time and cost savings for each task; the estimated time to build; and a suggested order, so you know what to build first.

Usually 5 to 10 working days after the discovery session.

Yes. The blueprint is yours to keep. You can use it with us, with another developer, or with your own team.

We do a limited number of blueprints each month, so that each one gets proper attention. We may say no if we do not think we can help your business. In that case, we explain why.

Yes. If you want a non-disclosure agreement before you share business details, we are happy to sign one. We treat everything you tell us as confidential either way.

It means up to 50% of the cost of the specific tasks we automate, not 50% of your whole business. For example, if your team spends £4,000 a month on invoice admin, an AI agent may cut that to about £2,000. Results depend on the task. Your blueprint shows the expected saving for your business, task by task.

We use your own numbers wherever possible: hours spent on each task per week; the hourly cost of the people doing it; how often errors happen, and what they cost; revenue lost to slow responses, such as leads that go cold; and the cost to build and run the AI agents. Where you do not have exact numbers, we use careful estimates and mark them clearly.

No. Savings are often the smallest part. We also show gains such as faster replies to customers, more leads followed up, fewer errors, and more capacity without new hires. We separate hard savings (money) from soft gains (speed and quality), so you can judge each one.

It depends on the project. Many simple automations pay back their cost within 3 to 6 months. The blueprint shows the expected payback time for each AI agent.

We measure results after launch and compare them to the blueprint. If an agent does not perform as expected, we adjust it.

Yes. You get the full calculation as a spreadsheet, with every input and assumption visible. You can change the numbers and see how the result changes.

A single AI agent usually takes 2 to 4 weeks from start to launch. A larger system with several agents takes 6 to 12 weeks. Your blueprint gives you a timeline before you commit.

1. We confirm the scope and timeline in writing. 2. We connect to your tools and build the first version. 3. You and your team test it with real examples. 4. We adjust it based on your feedback. 5. We launch it, often first with a small part of the work. 6. We train your team and hand over the documentation.

Yes, and we recommend it. Many clients start with one AI agent for their highest-value task. When it proves its value, they add more. This keeps your risk and your first investment low.

In most cases, yes. We connect to common tools such as Gmail, Outlook, Microsoft 365, Google Workspace, Xero, QuickBooks, HubSpot, Salesforce, Shopify and Slack, along with many more. If a tool has no standard connection, we can often still connect it. We check this during discovery.

About 1 to 2 hours per week from one main contact, plus some testing time near launch. We keep meetings short and focused.

Small changes are part of the normal process. If you want to add a new task or a big change, we agree the extra time with you first. We never add work without your approval.

Your data stays in your own systems wherever possible. Data that the AI agents must process is handled in UK data centres from trusted cloud providers. We tell you exactly where your data goes before the build starts.

No. We use business versions of AI services, which do not use your data to train their models. We check this for every service we use.

Only the people who build and support your system, and only when they need to. Access is protected and logged. All our team members sign confidentiality agreements.

Yes. We follow UK GDPR and the Data Protection Act 2018. We sign a data processing agreement with every client. We only process the personal data that a task needs, and we can help you update your privacy notice if your process changes.

No. Each agent can only use the tools and actions we give it. For example, an agent that drafts quotes cannot send payments. You decide which actions need human approval.

Every action is logged. You can see what the agent did, when, and why. We can also set up a simple dashboard or weekly report that shows the agent's work and results.

Yes, at any time. You can pause any agent instantly, and your team can take over the task by hand.

We design agents to know their limits. If a request is unclear or outside the rules, the agent passes it to a person on your team instead of guessing.

Most customers care about fast, correct answers. AI agents give faster replies at any time of day. You decide if the AI writes in your name, signs as an assistant, or only prepares drafts for your team. We match your tone of voice.

We help you. We can join a team meeting to explain what the AI does and does not do. Teams usually accept AI faster when they see it remove the work they like least. We also involve key team members early, so the system fits how they work.

Every build includes 30 days of free support after launch. During this time, we monitor the agents closely and fix any issues. After that, you can choose a paid support plan, or pay only when you need help.

We monitor your agents and get an alert when something fails. We aim to respond within 4 working hours for urgent issues. If an agent stops, the task goes back to your team, so nothing gets lost.

Software tools and AI models update often. On a support plan, we keep your agents working when your tools change. We also upgrade the AI model when a better option becomes available.

Yes, if we review them. We look at the agent's results, find where it hesitates or makes errors, and adjust its instructions and rules. On a support plan, we do this review every month or quarter.

Yes. Most clients start with one or two agents and add more over time. Your blueprint already shows the next best options, so new builds go faster.

You do. After final payment, you own the agents, the workflows and the documentation we build for you. Some parts may run on third-party tools, which stay under their own licences.

No. You get full documentation and access to everything we build. If you want to move to another provider or your own team, we help with the handover.

Because you own the system and the documentation, another developer can take it over. Your agents run on your accounts or on standard cloud services, not on a closed platform that only we control.

Yes. Support plans run month to month with 30 days' notice. Your agents keep running after you cancel. You only lose our monitoring and updates.