A powerful example of AI used thoughtfully to solve a real clinical bottleneck. MARS turns an hours-long specialist task into something usable in everyday stroke care, with the potential to improve support for hundreds of thousands of people living with aphasia in the UK alone.

The MARS Project is tackling a hidden gap in stroke rehabilitation. Around 350,000 people in the UK live with aphasia after stroke, yet most NHS speech and language therapists do not have the time or tools to analyse everyday talk properly. Only 30% routinely collect discourse samples, only 5% routinely transcribe them, and 95% rely on impressionistic judgement rather than detailed analysis. MARS uses programming and AI in combination to record, transcribe and analyse speech from people with aphasia, making this work practical in real clinical settings. The results are striking. The app can complete analysis in around 6 minutes compared with 8 hours 44 minutes for a trained human, a 99% time saving. Even when corrected by a therapist, it still cuts time by 84%. This means therapists can offer more personalised support, improve outcomes, and focus on helping stroke survivors rebuild the communication skills that matter most in daily life.

Vote now for this finalist to win the Tech4Good People’s Award and register to join the online ceremony on Friday 3 July to see if they win!