Personalised Learning Bot

Tender description

Cardiff and Vale College (CAVC) is seeking to appoint a supplier as part of an innovative project to develop a personalised learningassistant (bot).The project will involve building a personalised learning bot that will drive adaptive and contextual Learning and Teaching, drivenby advanced Artificial Intelligence (AI) and Deep Learning to help improve attainment in GCSE English and Maths.Anticipated core features and expectations include:• The teachers can be augmented with a self-learning Artificial Intelligence capable of creating a single knowledge from variouslearning sources while anticipating specific needs and personalising content.• Using machine learning and deep learning, the system will enable personalised self-service through a common canvas of proactiveengagement that is led by contextual tasks and activities, this will include embedding nudges to increase motivation andengagement.• The learning assistant will reduce the marking workload of staff, improve feedback to students, drive individual capability-basedlearning (which has been proven to have a huge impact on learners’ achievement), and provide a personalised learning path for theindividual.• The system will be driven by a voice ready, natural language typed exchange that will reduce the barrier to adoption whilemanifested within any other channel.• Whilst a teacher is marking a piece of work, the AI assistant will provide suggested content recommendation based on the marksand feedback of the teacher. The teacher will select the appropriate content for the learner to go alongside the marks and written oraudio feedback to the learner. This content is pulled in from existing sources, like the VLE, and is personalised to the learners needs,but saves time for the teacher in having to search for the content. It will also provide richer feedback for the learner as generallyfeedback just contains text, rather than content to support the areas of weakness.• The AI assistant can then follow up with the learner after they have received the feedback to find out if the content was useful, toensure that learning has taken place, and to offer further support.• The AI assistant can gamify and scan learners’ work for any dyslexia patterns before carrying out further diagnostics, which willnotify the necessary support staff of requirements for further ALN support. This will reduce the likelihood of learners who requireadditional support remaining undiagnosed, resulting in a more proactive approach to ALN.• The system will be bilingual using a language translation module.The project is being supported through the Welsh Government Digital 2030 capital grant funding. The project needs to be deliveredby May 2020.The College is inviting initial interested suppliers to register their expression of interest with a view to launching a formal invitationto tender process in January 2020. The maximum budget is £150,000, with a desirable requirement to exclude annual license feesfollowing development of the product.