UCL Centre for Digital Innovation


Psyrin: Leveraging AI for rapid assessment of serious mental illness through speech analysis

Startup Psyrin integrates clinial expertise with the power of AI to analyse speech patterns, resulting in the provision of a rapid and objective standard of assessment for serious mental illness.

Psyrin Presentation

12 September 2023

Psyrin is an innovative startup driven by its vision to create a harmonious clinical care system that integrates the expertise of clinicians with the power of AI, all to ensure the best possible patient outcomes in serious mental healthcare. To achieve this, they have introduced a groundbreaking solution that utilizes AI to analyse speech patterns, providing a rapid and objective standard of assessment for serious mental illness. In just 5 minutes of speech analysis, Psyrin's algorithms deliver valuable readouts, empowering mental health providers to assess early, monitor, and intervene effectively in cases of psychosis.

The challenge

Serious mental illnesses (SMI) are devastating – those living with SMIs are up to 20 times more likely to commit suicide (Zaheer et al. 2020) and have an 18-year reduction in lifespan compared to the general population (Chesney and Fazel 2014). These figures stem from a broken clinic care pathway that is reliant on reactive care, culminating in massive costs – an estimated £180 billion worldwide. Effort to solve this with the implementation of early, preventive interventions are constrained by existing diagnostic evaluations, which are lengthy, subjective and often inaccurate – a diagnosis of bipolar disorder (a prevalent SMI) takes on average 9.5 years (Bipolar UK, 2022). Consequently, thousands of undiagnosed and mismanaged cases impose burden on the economy, costing the UK over 6.5 billion every year.

The solution

Psyrin is introducing a new standard of clinical assessment for SMIs, with a particular focus on psychosis. With Psyrin, psychiatrists will save previous time and effort in the diagnosis of psychosis, as well as improve capacity by evaluating this condition at a much larger scale. Their proprietary machine learning algorithms analyse self-recorded speech samples (via their web application) to enable the early detection of psychotic disorders. The models were trained in a large proof-of-concept study, where we collected over 20,000 minutes of speech, and demonstrated preliminary accuracy measures of between 85% - 99% in several diagnostic classification tasks. With our product, clinics and clinicians will save time and costs, ensuring that preventive interventions at scale are the new standard of care.

Here’s how it works: a clinician sends a unique link to a potential service user. The user accesses this link to download the mobile application and follows a series of onscreen instructions, which take the user through a series of speech tasks, automatically recording their responses along the way. After completing these tasks, the user sees an immediate summary of their psychosis risk and suggested further resources. A more detailed breakdown is also sent to a clinician’s dashboard, where they can monitor and manage all service users.

CDI impact

Psyrin's mission revolves around expediting the adoption of preventive care for serious mental illnesses. This objective is only attainable through optimized and accurate assessments, forming the core of their central product. The impact of this much-needed shift is significant not only from a human perspective but also economically. Alongside substantial improvements in quality of life and relationships, Psyrin's innovation is poised to save billions in clinical system spending on suboptimal care provision annually.
From its inception, Psyrin's vision found a supportive ally in the CDI, aiding in translating their journey into a viable product roadmap. The CDI programme encouraged the swift development of a prototype through a combination of resources and pivotal advisory sessions. As they developed a prototype web application during the CDI program, Psyrin was able to showcase their actual technology to patients and clinicians, their ultimate customers. This demonstration has played a crucial role in establishing the desired research partnerships required to prove their clinical efficacy – nothing conveys their technology better than a working demonstration. Although there lies a substantial path ahead to achieve the intended impact scale, Psyrin has established a robust foundation – with the CDI playing an integral role in this progress.

Psyrin co-founders summarise their experience with the CDI:

“It was a privilege to work with Bruno (CDI Chief Technology Officer) – his guidance and hard work helped us overcome a key technical bottleneck in our ML timelines, and he introduced us to fundamental concepts in systems design.” - Julianna Oláh, CTO and Co-founder of Psyrin
“The CDI experience was transformational in our product journey – it has saved us immense amounts of time. The CDI’s intense focus on technical development is unique and truly a boost for any startup.” - Edwin Wong, CEO and Co-founder of Psyrin


  • Zaheer, J., Olfson, M., Mallia, E., Lam, J.S.H., de Oliveira, C., Rudoler, D., Carvalho, A.F., Jacob, B.J., Juda, A. and Kurdyak, P. (2020). Predictors of suicide at time of diagnosis in schizophrenia spectrum disorder: A 20-year total population study in Ontario, Canada. Schizophrenia Research, [online] 222, pp.382–388.
  • Chesney, E., Goodwin, G.M. and Fazel, S. (2014). Risks of all-cause and Suicide Mortality in Mental disorders: a meta-review. World Psychiatry, [online] 13(2), pp.153–160.
  • Bipolar UK. (2022). The Findings of the Bipolar Commission. [online] Available at: https://www.bipolaruk.org/bipolar-commission-findings#:~:text=More%20tha... [Accessed 1 Sep. 2023].