We’re delighted to present our finalists for the category of “Best AI scribe solution”.
Alder Hey Children’s NHS Foundation Trust: Alder Hey and Lyrebird Health deliver trust wide AVT AI scribe deployment
Overview: Alder Hey partnered with Lyrebird Health to deploy their AI scribe across the trust. Using ambient voice technology and natural language processing, Lyrebird automates documentation during consultations, reducing administrative burden.
What happened? Lyrebird listens to the clinical consultation and automatically generates structured clinical notes and documentation within the clinician’s existing workflow. The project was clinically led and designed. Pre-implementation analysis identified documentation pain points and ensured the AI scribe addressed real needs. Initial deployment focused on ED and outpatients, followed by community and neurodevelopmental services. Key implementation features included integration with AlderCare EPR, compliance with NHS England AVT AI guidance, a phased rollout approach, and tailored training, user support and floor-walking. By December 2025 over 900 staff have been onboarded, with more than 2,000 consultations per week supported by Lyrebird. User surveys from an initial cohort of 56 users reported eight minutes per patient per clinic compared to traditional methods. 95 percent of clinicians surveyed would recommend Lyrebird to a colleague, and 100 percent reported improved ability to focus on patient interactions. A 7.2 percent uplift in clinical note and document quality was observed.
ScienceSoft: Human-like, real-time AI scheduler powered by Amazon Nova Sonic by ScienceSoft
Overview: ScienceSoft developed an AI voice scheduler for healthcare, enabling secure patient appointment management via voice. The solution ensures privacy, automates scheduling, and integrates seamlessly with existing systems. It enhances efficiency, reduces administrative workload, and delivers an accessible patient experience.
What happened? The solution delivers natural, real-time, human-like voice interactions that integrate with core healthcare systems. It combines generative AI, secure cloud architecture, and healthcare data interoperability for automated patient engagement and operational efficiency in clinical scheduling. Amazon’s Nova Sonic speech-to-speech model with the LiveKit Media Server enables true bidirectional streaming and real-time conversational dynamics through a single API. This allows the scheduler to listen and respond without noticeable latency, adapt to speech patterns, tone, and natural pauses, and interact directly with backend systems to check provider availability, verify identities, and record scheduling decisions. The solution supports both inbound and outbound calls. Patients can call to book or modify appointments, while the system can proactively reach out to confirm or reschedule. Deployment resulted in a 40 percent reduction in appointment booking time, a 50 percent reduction in operational costs, and a 70 percent increase in scheduling throughput per hour compared to human agents.
Modality Partnership x Heidi Health: Releasing time to care: Scaling Heidi Health’s Ambient AI scribe across Modality Partnership
Overview: Modality has embedded Heidi Health across 50+ sites and 200+ clinicians. Over 1.2 million minutes of consultations have been transcribed, cutting documentation time by more than half, improving note quality, reducing burnout and unlocking hundreds of clinical hours every month.
What happened? A 25-day pilot with 47 GPs, 2,879 consultations and 24,156 minutes of transcribed time more than halved documentation time. Modality moved rapidly to scale rollout to 190+ Modality clinicians (now >200) across GPs, nurses, pharmacists and PAs. Over 1.2 million minutes of consultations have been transcribed to date, with 17,000+ consultations auto-transcribed per month, saving 530 clinical hours each month. Heidi has delivered major, sustained efficiency gains, with a 51 percent reduction in documentation time during consultations, and a 61 percent reduction in documentation completed after hours. GPs typically reclaim 1.5–2.5 hours per day, equivalent to around 12.5 hours per week – an extra clinical session. A 58 percent reduction in documentation-related stress was observed, along with a 45 percent improvement in satisfaction with work–life balance. Also of note is a 38 percent improvement in clinician satisfaction with note quality, with a 33 percent increase in confidence that notes accurately reflect the consultation. Patients describe feeling “more listened to” and more connected with their GP.
X-on Health: Surgery Intellect powered by TORTUS: The NHS-proven AI scribe
Overview: Surgery Intellect powered by TORTUS improves documentation quality, reducing consultation time and freeing clinicians for patient care. With telephony integration, real-time monitoring and NHS evidence, it scales safely across primary care to transform workflows, reduce burnout and unlock capacity.
What happened? By combining TORTUS’ Ambient scribe with X-on Health’s Surgery Connect telephony platform, the solution supports both in-person and telephone consultations. Because Surgery Connect handles millions of telephone consultations, Surgery Intellect can automatically scribe both remote and face-to-face encounters. TORTUS has developed an active monitoring layer that tracks scribe performance, hallucination rates and workflow behaviour. A study coordinated by Great Ormond Street Hospital (GOSH), evaluated TORTUS across nine NHS organisations in London. More than 17,000 real patient encounters were analysed. The results showed a 23.5 percent increase in patient interaction time, an 8.2 percent reduction in appointment length, a 35 percent reduction in clinicians feeling overwhelmed by notetaking, and 92 percent patient/carer acceptance. In A&E, 13.4 percent more patients were seen per shift. 100 percent of AI-assisted notes were rated “good” or “very good”, versus 43 percent for typed notes and documentation quality more than doubled.
Heidi and The Dudley Group NHS Foundation Trust: Heidi AI scribe pilot at The Dudley Group NHS Foundation Trust
Overview: Dudley Group NHS Foundation Trust (DGFT) has a 30,000+ outpatient letter backlog across 42 specialties. Clinics are paper based and slow letter workflows delay patient and GP communication, impacting care and patient experience. Coding delays risk financial loss. The trust piloted the AVT AI scribe since January 2025.
What happened? DGFT implemented the Heidi AI Scribe in Same Day Emergency Care (SDEC), SDEC outpatients, rheumatology outpatients and from April to July 2025 the Trust piloted clinical coding for the first time. A proof of concept for integration with the Sunrise EPR will enable real-time transcription during consultations, with immediate generation of clinical letters. Clinicians could review notes at the point of care, reducing post-clinic administrative work. In SDEC, documentation time per patient fell from 8.6 to 2.5 minutes, freeing 221 clinical hours per month for direct patient care. In rheumatology, the backlog of over 2,700 clinical letters was reduced to fewer than 200 within five months. Turnaround times improved from six months to eight days. Patients consistently reported improved consultations, with over 90 percent feeling more listened to. Documentation omissions decreased by 55 percent, coding queries dropped 15 percent, throughput tripled, completion times decreased 45 percent, and errors fell 90 percent.
The Royal Wolverhampton NHS Trust: Delivering safe, scalable impact at RWT through governance-led adoption of ambient voice technology
Overview: The Royal Wolverhampton NHS Trust partnered with CLEARnotes to safely implement Ambient Voice Technology in preoperative clinics. Using a governance-led, clinician-driven approach, the pilot improved productivity by 25 percent, reduced administrative burden and costs, and enhanced patient communication.
What happened? Partnering with CLEARnotes, RWT piloted AVT in consultant-led preoperative anaesthetic clinics. The pilot was clinically led, and the approach included comprehensive site readiness, pathway mapping, prompt engineering, testing, evaluation, joint training and support structures. Robust IT, cyber security and information governance assurance was fully aligned with trust and national standards, and clinical safety compliance adhered to DCB0129 and DCB0160 standards. Results included a 25 percent increase in patients seen in one preoperative clinic, equivalent to £87,360 potential annual income if scaled across ten consultant clinics over 42 weeks. There was a 75 percent reduction in letter turnaround time (6 days to 1.5 days), a 63 percent reduction in administrative time, releasing 46 minutes per clinic, and a 67.9 percent reduction in additional waiting list initiative costs, equating to £38k annualised savings if sustained. Letters are now produced in approximately ten seconds, accelerating onward care. The model is being piloted in nine additional outpatient departments.
InventAsia Limited T/As Prescribe Digital: AISA® – Ambient AI that automates the journey from clinic room to final distribution
Overview: AISA® Ambient AI scribe delivers end-to-end clinical documentation. It captures consultations, distinguishes clinician intent, generates structured outputs, and supports full clinician validation. Integrated with EPRs, AISA® reduces documentation time by up to 70 percent while improving safety, accuracy, and experience.
What happened? AISA® distinguishes between patient narrative, clinician reasoning, and clinician instruction. This allows a single consultation to produce multiple targeted outputs such as clinical notes, GP letters, and patient letters. Once a clinician has reviewed and approved the generated documentation, AISA® can write structured data back into the EPR at the field level. When integrated with a capable EPR, clinician instructions can be used to initiate order communications or trigger automated actions such as referrals or follow-up tasks. Low-confidence content is flagged, and all edits are logged with user and timestamp, supporting full medico-legal traceability. Early NHS evaluations show AISA® reduces documentation time by 60–70 percent, saving up to 15 minutes per consultation. This equates to approximately 256 hours, or 32 working days, returned to patient care per clinician each year. Organisations report faster turnaround of outpatient correspondence, reduced transcription dependency, improved EPR data quality, and improvements in clinician wellbeing.
