NHS Grampian partners with Corti to bring ambient AI documentation to frontline ward care
.png)
TL;DR
Corti is expanding our NHS partnership into Scotland, working with NHS Grampian and the NHSG Innovation Hub to bring ambient clinical documentation to Inverurie Hospital's Donbank Ward.
Documentation time is down nearly 70%. Nurses describe feeling more present. Handover quality between shifts has improved. And the team is tackling one of the most distinct dialects in Scotland, teaching the model Doric in real time.
This piece covers what changed at the ward, why handovers improved, how local dialect adaptation works in practice, and what it points to for ambient AI across the NHS.
Corti is partnering with NHS Grampian to bring ambient clinical documentation to Scotland for the first time, marking the latest expansion of Corti's work with NHS trusts across the UK.
The NHS faces a significant and well-documented administration burden. Clinicians spend hours each week reconstructing encounters after the fact, time that comes directly out of patient care. Ambient voice technology is one of the most promising tools available to address that, and this pilot is a live test of what it can deliver in practice.
The pilot, led by NHS Grampian's Innovation Hub and supported by Aberdeenshire Health and Social Care Partnership, is running at Inverurie Hospital's Donbank Ward in the northeast of Scotland. It puts Corti's ambient voice technology in the hands of ward nurses, with one goal: reduce the time spent on documentation and give that time back to patient care.
Early results from the ward are direct. Nurses report that documentation time has dropped nearly 70% and describe feeling more present in conversations, and the quality of information handed over between shifts has improved. Senior charge nurse Katie Anderson, who is leading the pilot, has called it a "massive difference."
There is also a distinctly local dimension to this deployment. The northeast of Scotland is home to Doric, a distinct Scots dialect, and the Donbank Ward team has taken on the work of teaching the model to understand it. Their corrections feed back into how the system learns, making this pilot as much about adaptation as adoption.
What follows is what the ward has found.
What changes when documentation happens in the background
Most nurses know what post-encounter documentation feels like. The patient is settled, the conversation is over, and then begins the work of reconstructing what was said, what was noticed, what mattered. Details compress. The emotional texture of a conversation gets flattened into observations. Notes become a record of tasks rather than a record of a person.
The question NHS Grampian's Innovation Hub set out to answer was straightforward: if ambient voice technology could handle that reconstruction in real time, what would nurses do with the recovered attention?
The answer, it turns out, is listen.
Corti's ambient documentation technology works by listening to conversations between clinical staff, patients, and families, then generating structured clinical notes that nurses review and edit before they're finalized. There's no dictation step, no post-encounter write-up. The documentation happens alongside the care.
For the nurses at Donbank Ward, the most immediate change was cognitive. Senior charge nurse Katie Anderson, clinical lead at the ward, described the shift clearly: "We used to have to try and remember what was said during observations and record this in notes later. Using the tech frees up mind space so we can have a really good conversation."
That freed attention changed the quality of interactions. Patients began sharing things they hadn't before: personal reflections, emotional concerns, things that don't always surface in a documentation-forward encounter. Staff reported that some patients, and their families, engaged differently once the dynamic shifted away from note-taking.
One patient commented that the transcript captured what she was actually feeling, not just a clinical summary of what had happened.
Anderson added: "Staff have been giving really good feedback about the meaningful conversations which using the technology has prompted them to have. Before, documentation could be very task-oriented and based more on observations rather than discussions. It's transforming how we reflect on patient care."
The handover effect
One of the less obvious benefits has been at shift handover. When notes are built from the conversation rather than reconstructed afterward, the depth and accuracy of what gets passed between teams improves. Nurses at Donbank Ward have reported that the quality of information handed over between shifts has increased, which has practical downstream consequences for continuity of care.
This is one of the things Corti's technology is built to address. Clinical speech-to-text is not a solved problem. The vocabulary is specialist, the acoustic conditions are difficult, and the stakes for accuracy are high. Corti's models are trained specifically on clinical language, structured to capture the kind of detail that matters in a healthcare context rather than producing a general transcription. The goal is notes that read like a clinician wrote them, because they reflect what a clinician said.
Teaching the model something new
What has emerged at Donbank Ward is also a useful illustration of how this kind of system actually improves. The northeast of Scotland is home to Doric, a distinct Scots dialect, and the ward team has taken on the work of adapting the model to recognize it. Nurses edit transcripts in real time, and those corrections feed back into how the model learns.
What began as a practical challenge has become something the team has taken ownership of. Anderson described it directly: "We've had lots of laughs with the Doric accent. It can confuse the system at times, but it's improving every day, and staff can edit live so the model learns from us."
This kind of local adaptation is part of how Corti thinks about deployment. Florian Schwiecker, Chief Partnership Officer at Corti, visited the ward earlier this year. "What struck me visiting Donbank Ward was how much ownership Katie's team have taken over the technology. They're teaching it Doric, they're shaping how it learns and they're gaining more quality time with patients because of it. That's why Corti exists – to provide specialised AI that understands how care actually happens, and where support is genuinely needed."
The team members who were most skeptical at the outset are now among the most engaged. That shift is worth noting. Ambient documentation tools often face adoption resistance in clinical settings, where the margin for error is low and trust in new technology is earned slowly. The ward's experience points to something important: when the technology demonstrably gives time back rather than adding process, the calculus changes.
What this points to
The administrative burden on NHS staff is not a marginal problem. It shapes how much time clinicians have with patients, how well information transfers between teams, and ultimately how care is experienced on both sides of the encounter.
Rebecca Thomson, Project Support Officer for NHS Grampian's Innovation team, framed it plainly: "By recording and automatically summarising conversations with patients, the technology enables staff to deliver more person-centred care while maintaining strong clinical oversight and patient safety. Corti has been developed specifically with healthcare in mind, and we are excited to see the potential benefit the system can provide to our teams across Grampian."
The findings from Donbank Ward will inform whether this approach can scale across NHS Grampian. But the more immediate signal from the ward is less about scale and more about what happens to care quality when the burden of documentation is redistributed. Nurses have more capacity to be present. Patients share more. The record of an encounter starts to reflect the encounter itself.
That's the version of clinical documentation worth building toward.
Join our mission
We believe everyone should have access to medical expertise, no matter where they are.


