News
From teaching challenge to potential spinout: Queen Mary AI assessment project takes next step
Faculty of Science and Engineering13 August 2026
An artificial intelligence project developed at Queen Mary University of London to help academics provide faster, more detailed feedback to students has been selected for a national spinout programme.
EduGrade AI, an educator-controlled assessment and feedback platform, has been selected for the Innovate UK ICURe Exploit Business and Spinout Readiness programme, following its successful completion of the ICURe Explore programme.
The project is now progressing towards spinout formation through Queen Mary Innovation, marking an important step in taking the project beyond the university and towards wider use across higher education.
Unlike many university spinouts, which originate in research laboratories, EduGrade AI began with a practical challenge in teaching: how can academics provide students with high-quality, timely feedback at scale, while ensuring that academic judgement remains with the educator?
EduGrade AI is designed around an educator-controlled, human-in-the-loop approach. Academics define their own assessment criteria and rubrics, while the platform uses AI to generate draft, rubric-aligned feedback and indicative assessment outputs. Educators review and edit the outputs and retain responsibility for all final academic decisions.
To date, EduGrade AI has been applied to more than 1,000 student submissions across different assessment contexts. Evaluation to date has demonstrated approximately a 60 percent reduction in educator marking time, while 86 percent of students surveyed reported that the feedback was helpful.
The project received £35,000 through Innovate UK's ICURe Explore programme, which supported market discovery and exploration of its potential beyond Queen Mary. Its progression to ICURe Exploit provides further business and spinout readiness support as the project explores its potential as a scalable commercial venture.
For the team behind EduGrade AI, the project also raises a wider question about how universities identify and develop innovations that emerge not from a laboratory, but directly from teaching and learning.
"EduGrade AI started from a practical challenge in higher education: how we can provide students with high-quality, timely feedback at scale without removing academic judgement from the assessment process. Progressing to ICURe Exploit is an important step towards translating this work from educational innovation into something that could benefit universities much more widely." said Dr. Deepshikha.
The project's approach is centred on using AI to support academics rather than replace them.
"AI should give academics their time back, not take their judgement away. Every decision that matters still sits with the educator," the team said.
EduGrade AI has also received wider recognition and support through initiatives including the Google Higher Education Faculty AI Fellowship, OpenAI's Professors Teaching with AI series, Queen Mary's President and Principal's prize for Educational Excellence, Innovate UK ICURe, MassChallenge UK, Microsoft for Startups and Google for Startups Cloud.
The move towards spinout formation reflects Queen Mary's wider ambition to translate ideas developed through its research and education into practical innovations with potential benefits beyond the university.
This project was developed and led by Dr. Deepshikha from the School of Engineering and Materials Science at Queen Mary University of London. You can learn more about her work and publications via the following Queen Mary Academy webpage.
Updated by: Laura Shepherd
