thymia is delighted to have partnered with Frome Medical Practice, an NHS GP practice, leveraging our voice-driven mental health biomarkers to create a tailored solution to enhance their social prescribing programme. Social prescribing is a key NHS initiative and cornerstone of the NHS Long Term Plan, which is committed to building infrastructure for social prescribing in primary care across England. As part of the largest national investment in social prescribing by any health system worldwide, it connects them to community activities, groups, and services that address the practical, social and emotional needs affecting their health and well-being. Through this approach, healthcare professionals refer patients to dedicated link workers who provide personalised support plans based on what matters most to the individual. Our collaborative partnership with Frome Medical Practice strengthens this vital NHS priority by using our AI technology to better identify patient needs and increase engagement with community resources. Our collaboration provides meaningful data through personal well-being assessments that help tailor social prescribing interventions for patients with long-term conditions, mental health needs, social isolation or complex needs—precisely the populations that NHS England has identified as benefiting most from this whole-population approach to healthcare.
Addressing real-world healthcare challenges
Frome Medical Practice and Health Connections Mendip (a specialist social prescribing and health coaching team) are advancing the social prescribing agenda in the UK by making this approach to treatment more accessible and widely used. Despite having many valuable initiatives available locally, these resources were being underutilised by target patient groups with challenges in long-term engagement.
This engagement gap is common in healthcare settings: patients often need additional support to begin and continue using available well-being services. Our work with Frome Medical Practice shows that people engage more consistently with health programmes when they can clearly see value and measure their progress over time.
The partners at Frome Medical Practice identified an opportunity to use thymia's technology to encourage more patients to access Health Connections services and maintain engagement with available initiatives. They selected our technology for its ability to analyse multiple wellness parameters while integrating smoothly with their existing workflows, helping connect patients to appropriate community resources.
A new approach to measuring well-being
Our voice-based technology transforms subjective wellness experiences into objective, measurable data, offering patients an engaging and insightful alternative to traditional questionnaires. Unlike standard self-report measures, which we know can be inherently inaccurate and unreliable due to their subjective nature, our approach better tracks well-being over time through more objective means.
Traditional wellness assessments often rely on self-report questionnaires that struggle with accuracy and consistency problems. Responses can be significantly influenced by mood at the time of completion, memory limitations, and tendencies to answer in socially desirable ways rather than truthfully. Additionally, these methods can feel burdensome and impersonal when used repeatedly and typically don't capture the detailed information needed to track well-being effectively over time.
Our approach offers a different solution. We built a tailored wellness webapp for Frome Medical Practice that provides patients with a well-being companion through engaging voice-based check-ins. By analysing patterns in voice, our technology transforms subjective experiences into objective, measurable data points. Patients can monitor their mental stamina, relaxation and energy levels independently and remotely. A single brief check-in—using short speaking activities—gives users insights into multiple aspects of their well-being, with visual tools that help them track changes over time.
This approach helps patients identify meaningful patterns. For example, some users discover that while low energy initially makes them reluctant to join social activities, participation actually increases their energy levels afterwards. These insights help motivate continued engagement with well-being programmes and build a foundation for better psychological well-being and healthcare outcomes.
We designed the user journey specifically for Frome members and patients, with built-in flexibility to adapt as their needs evolve.
Benefits for healthcare providers
In an increasingly time-pressured NHS environment where clinicians struggle to gather comprehensive patient well-being data between limited appointments, our technology offers a solution that maximises clinical effectiveness without adding to workload burdens. GPs can prescribe the tool to patients, enabling continuous monitoring of well-being between appointments. This provides clinicians with objective data about patient well-being over time, similar to how they might monitor other physical health markers, supporting more informed follow-up conversations and allowing for more precisely targeted interventions.
From a clinician or healthcare worker's perspective, in-depth tracking of different aspects of mental well-being can help give a more detailed view of what's going on in clients' real lives between appointments — acting as a springboard for discussion to identify what's working and what might need adjustment to keep users on track.
Technical implementation: How our technology works
The integration of our bespoke wellness solution for Frome Medical Practice delivers on several parameters. Our implementation enables patients to conduct regular well-being assessments through structured check-ins, generating longitudinal data that also encourages self-monitoring capabilities. The system streamlines access pathways to community resources as next steps, significantly improving resource awareness within their social prescribing framework.
Our ability to provide this transformative solution stems from our unparalleled expertise in voice-based biomarker technology and our uniquely wide and deep proprietary mental health dataset. This means that our models are built from datasets that are both large and diverse in terms of people represented (10s of thousands of voices, across ages, genders, cultures, languages/accents, and health statuses), and that we have specific expertise in learning to recognise signals of well-being in the voices of specific individuals we have followed over long periods time.
What sets our approach apart is the transformation of traditionally subjective health assessment into objective, measurable insights. Until now, mental well-being evaluation has largely relied on patients' ability to accurately report and recall their experiences. Our deeptech AI technology detects subtle patterns in voice that reliably indicate mental state, alongside other biomarkers that are traditionally complex to detect. This objectivity is crucial for healthcare settings where consistent, comparable measurements are needed to track progress effectively and guide treatment decisions. By removing the biases inherent in self-reporting, we provide a more reliable foundation for understanding and supporting overall health and well-being, creating new possibilities for measurement across multiple domains previously limited to subjective assessment.
Our solution's efficacy relies on two distinct types of models:
Cross-sectional models: These benchmark individual patients against demographically matched cohorts, accounting for variables including age, gender, language, accent, ethnicity, and cultural background. This establishes baseline normalisation parameters and provides immediate contextualised assessment.
Longitudinal models: Unique in the healthcare AI sector, these models build personalised assessment frameworks for returning users. After three consecutive data collection points, the system adapts to calibrate individual speech patterns and behavioural markers, offering increasingly precise analysis with continued use. This technological breakthrough allows us to capture subtle changes in mental state that traditional assessments simply cannot detect.
Importantly, both model types incorporate research-based knowledge about how different aspects of psychological well-being interact and change over time. Our models have been thoroughly tested for psychological validity to ensure each measurement reflects specific aspects of well-being rather than unrelated factors. This ensures that each score provides independent information without being influenced by other physical or mental health-related factors.
This technical approach enables both immediate insights and progressive personalisation that conventional systems cannot achieve.
User perspectives:
Patients often struggle with maintaining consistent engagement in well-being programmes due to a lack of motivation and limited self-awareness, yet our user feedback demonstrates how technology can bridge this critical gap. Through our extensive user testing, we've gathered powerful feedback from webapp users that underscores the impact of our technology:
"It helps me understand my mental state" -user
"Using the daily check reminded me to check in on myself" -user
"thymia offered me an insight, understanding, and a chance to reflect" -user
"It gave me a reason not just to get out of bed but to try, it felt like someone was listening and there for us" -user
"I like the regularity of using thymia, and the chance to reflect on how I am feeling" -user
These testimonials show how our solution helps build self-awareness and motivation—key factors in maintaining engagement with well-being programmes and addressing the core challenge Frome Medical Practice identified.
Beyond healthcare applications
The Frome implementation showcases one application of thymia's technology, demonstrating the significant potential of our AI models to enhance healthcare delivery across multiple sectors.
This deployment demonstrates how our AI technology seamlessly integrates into existing NHS systems, generating valuable data for continuous improvement. By providing patients with personalised insights and enabling healthcare practitioners to make data-driven decisions, we're helping to improve the patient experience within the NHS.
By providing patients with accessible insights into their mental well-being and connecting them with appropriate community support, our technology helps bridge the gap between available services and the people who need them. For healthcare providers, it offers a data-informed view of patient progress that supports more effective care.
Similar engagement challenges exist in other contexts such as Employee Assistance Programmes (EAPs) or employers who want to track the mental well-being of people engaging in specific programmes. For healthcare specifically, we continue to develop our capabilities, including models for measuring clinically significant depression, anxiety, and post-traumatic stress symptoms.
As we expand our work, our focus remains unwavering: developing versatile AI solutions that deliver practical, measurable outcomes while generating rich data for continuous improvement. The successful integration with Frome Medical Practice demonstrates how technology can enhance healthcare outcomes and improve support systems through the thoughtful application of AI.