Memory Lane Games, in collaboration with experts from SAP and EY, has developed a generative AI prototype that builds personalized memory games from life stories, family memories and photographs to assist people living with dementia and Alzheimer’s disease. The initiative aims to offer a more personal game experience that encourages reminiscence and connection, while also easing everyday tasks for family members and caregivers.
How the prototype works
The existing Memory Lane Games digital platform already offers simple, quiz-based games designed to start conversations and evoke positive memories. The new AI prototype extends that approach by recommending topics, images and multiple-choice questions based on an individual’s cultural background and life history.
To create personalized content, the system uses a profile that can include details such as age, native language, previous places of residence, occupation, favorite foods, hobbies, travels and pets. These specifics let the system generate questions and visual content that are tailored to the person rather than generic.
Why personal memories matter
Developers found that for many people living with dementia the important anchors are not broad topics but the small details tied to personal stories: a former home, a favorite hiking spot, a family recipe or an old pet can all trigger conversation.
The prototype suggests topics tailored to the individual profile that can help recall positive memories and bring patients, family members and caregivers closer together. The aim is therefore more than gameplay: it is to create shared experiences with genuine emotional value.
Development collaboration and pro bono support
The work was carried out within the MovingWorlds Scaling AI for Good program, which provides professional assistance, pro bono consulting, workshops and partner access for social enterprises with the involvement of SAP, EY and Microsoft.
In the Memory Lane Games project, SAP developers and EY technology experts explored how to combine generative AI with visual information and context derived from life stories. They tested multiple approaches, including generating questions first and then finding images, or starting from images and generating related content. Based on their experiments, the image-first approach proved more effective.
Current status and possible future features
The prototype is currently in a testing phase. Future developments may allow family members and caregivers to upload personal photographs; with proper consent and data handling, metadata linked to those image files could be used to create personalized questions. The development team also sees opportunities for voice-based and multilingual solutions and AI assistants tailored to the individual needs of people living with dementia.
The value of technology lies in human impact
Hidvégi Péter, managing director of SAP Hungary, noted that corporate conversations about AI often focus on efficiency and automation, but this project demonstrates how technology can create value on a different level: by making personal stories accessible again and helping to produce moments of connection that matter to people and families.
SAP has supported social initiatives for more than a decade by involving employee expertise. Such collaborations provide social enterprises with knowledge and technological experience that would be harder to obtain on their own, while SAP employees deepen their practical understanding of AI and other technologies by working on real-world challenges.
The Memory Lane Games development illustrates how artificial intelligence can become a tool for social innovation, with an emphasis on making personal memories more accessible and enabling enjoyable, meaningful human connection rather than applying technology as an end in itself.
