NHS Intent capture design history

We’re starting a new piece of work exploring how AI could help people find the right NHS service or support, first time.

The problem we’re looking at is familiar but difficult. People often come to the NHS with a need, a worry, a symptom or a question, but they do not necessarily know where that need fits within the structure of the NHS.

They may not know whether they need 111 online, a GP, pharmacy, urgent treatment centre, A&E, sexual health services, or something else entirely.

Although the NHS is organised by clinical pathways, services, and responsibilities, individuals typically perceive and navigate the system based on their own personal experiences and needs.

This work is about closing that gap.

What we are exploring

We are exploring an AI-powered navigation service that allows people to describe what they need in their own words, without first needing to choose the right service, topic or category.

The aim is not to create a clinical conversation or replace existing NHS services. The aim is to help people get to the right place more quickly, more safely and with less friction.

That might mean helping someone:

• understand whether they need urgent help
• find the most appropriate NHS service
• move into an existing pathway, such as 111 online
• find trusted health information, such as Health A to Z
• find local services using NHS services near you
• understand what the Intent capture service can and cannot do
• leave the service safely if it is not helping them

Why this matters

At the moment, people often need to understand the structure of the NHS before they can successfully use it.

They may need to know the difference between 111 online, NHS 111 by phone, a GP, pharmacy, A&E, urgent treatment centres, service finders, Health A to Z content and local services.

That is a lot to expect from someone who may be worried, unwell, caring for someone else or under pressure.

Good navigation should not require people to already know the map.

Our design challenge

The central design challenge is not simply:

How might we use AI?

It is:

How might we help people find the right NHS help, first time, without needing to understand the structure of the NHS?

That raises a wider set of design questions:

• Where should this service appear?
• How should people enter it?
• How do we explain what it can and cannot do?
• How do we make the use of AI transparent?
• How do we identify emergencies quickly?
• How do we avoid creating loops between services?
• How do we connect safely with existing NHS pathways?
• How do we measure whether the outcome was useful?
• How do we make sure people always have a non-AI route?

Our design principles

As we begin, we are trying to hold a few principles in mind.

The service should be useful before it is clever

It should reduce effort, not add another layer of complexity.

The value of Intent capture is not that it uses AI. The value is whether it helps people get to the right place more easily than they can today.

The service should show its seams

People should understand when they are using AI, what it is doing, what it is not doing, and where responsibility sits.

This is especially important in health, where people may assume that a digital conversation is clinical, personalised or being monitored by a professional.

The service should be safe by design

Emergency needs must be identified quickly, clinical boundaries must be clear, and people must not be left thinking someone will follow up if that is not the case.

Where someone needs urgent or emergency help, the service must make that clear and direct them appropriately.

The service should work with the NHS, not around it

This work needs to connect into existing services, standards, content, clinical safety processes and operational realities.

It should support services such as 111 online, NHS services near you, Health A to Z and, where relevant, professional tools such as NHS Service Finder.

The service should be inclusive from the start

People describe health needs in different ways, with different levels of confidence, literacy, language and context.

The service needs to account for that.

It should not rely on people already knowing clinical terms, NHS service names or how the system is organised.

What this design history is for

This design history will capture how the work develops.

A design history helps teams record design decisions and create a permanent record of how a service has changed over time. The Department for Education describes a design history as a public record of what has been designed and why, helping teams tell the story of how a service has developed.

This design history will record:

• the questions we ask
• the assumptions we test
• the decisions we make
• the risks and constraints we identify
• the things we learn from users, stakeholders and subject matter experts
• where our thinking changes

We expect the work to cover interaction design, content design, service design, user research, product thinking, clinical safety, governance, policy and technical exploration.

Some of the questions will be practical. Some will be strategic. Some will be uncomfortable.

That is the point of recording the work as we go.

Intent capture is not just a new interface pattern. It asks bigger questions about how people find help, how the NHS presents itself, how trust is built, and how we design responsibly when the technology is still emerging.

This design history is a place to make that thinking visible.