How the UK's biggest care home charity transformed policy access with AI
MHA supports thousands of residents, families, and staff across 75 care homes, 59 retirement living communities, and 43 community groups. Finding the right policy information quickly isn't just a convenience – in a care environment, it matters.
The impact at a glance
Down from 10 minutes.
Searchable, retrievable and readable in plain language.
Scaling across MHA's full estate.
The challenge
MHA is one of the UK's largest and most trusted care home providers – supporting thousands of residents, their families, and the staff who look after them across 75 care homes, 59 retirement living communities, and 43 community groups.
Like many large organisations, MHA had accumulated over 900 internal policies, documents and guides that staff needed to navigate daily. Finding the right information quickly wasn't just a convenience issue – in a care environment, it matters. The question was whether AI could genuinely help, or whether it would just add another layer of complexity to already demanding jobs.
What we did
Brilliant Noise partnered with MHA to design and build 'iSPoC' – an AI-powered policy assistant that acts as an internal single point of contact for policy queries. Staff can search, interpret and retrieve policy information instantly, in plain language.
We started by building a clean, query-friendly index of MHA's policies – the foundation that makes fast, accurate retrieval possible. From there, we designed an easy-to-use interface and worked closely with MHA leadership throughout, shaping the tool in response to real feedback rather than assumptions.
The pilot ran across eight care homes. Early response times of 40 seconds dropped to just 4 seconds. We added multilingual functionality – a meaningful step given the diversity of MHA's workforce.
But the work didn't stop there. Pilot feedback directly shaped the next phase of development. iSPoC has since moved into MHA's own secure environment – giving greater confidence around data handling, access control and long-term sustainability, and making it much easier to scale responsibly across the organisation. We've doubled the number of policies we're testing, the interface has been overhauled for usability, and new features including secure login and conversation history have been introduced. Response accuracy has been tightened through more rigorous testing, and in-tool feedback has been simplified so staff can share their thoughts without friction.
The impact
The moment that crystallised the project's value came during a board meeting. A policy question that had previously stalled the room for 10 minutes was answered by iSPoC in under a second.
That kind of clarity changes things. Staff report feeling more confident and more capable – not just faster at finding answers, but less anxious about getting them wrong. The project has shifted how people at MHA think about AI: not as a distant technology topic, but as something that already makes their working day better.
Leadership sees clear potential to scale – across MHA and as a model for the wider care sector.
The results
- Policy search time: from 10 minutes to under a second.
- Faster decisions: instant access to accurate information means care home staff can act with confidence rather than waiting.
- Cultural shift: frontline staff report higher confidence and genuine curiosity about what AI could do next.
"This is a really important and innovative initiative for MHA. There was real appreciation of the value of the intent, the opportunity, and the sheer amount of work contributed. It demonstrates the genuine innovative capability and culture that's being fostered and developed across our organisation."Jonathan Mace Director of Communities Charitable Giving and Marketing