• CiCi is currently in a first beta version release.
• We have many plans for further development.
• This web page provides a list of some of those plans – both to improve the technologies and interface – and to provide more content.
• We welcome reports of bugs in our beta release and feature requests for future versions.
Improvements in CICi understanding of user utterances
It can be frustrating when CiCi does not understand what you are saying. CiCi incorporates Artificial Intelligence technologies, in particular Natural Language Understanding (NLU). As an ongoing activity we are using Machine Learning to improve CiCi’s understanding, and the more CiCi is used the better the bot should become.
Improved understanding of job titles
Our data around the content of different occupations comes from official UK job titles and information round job vacancies from outside suppliers. Unfortunately there is no standard for how job titles and jobs are described, although this may emerge in the future. We can use AI to help CiCi understand what people are asking and use autocompletion of job titles to match standard descriptors. But there are limitations to what we can do with data accessed by API from external sources.
Better local course data
At the moment our course information is based on a spreadsheet published by the governments Skills Funding Agency (SFA), updated monthly. This is limited to courses which receive central funding from the agency. We are currently working with colleagues from our partner cities, in Bristol. Derby and Newcastle, to access local course information, for instance from the Adult Education Service.
For finding courses we are currently searching course descriptions from the 44000 courses listed in the ESFA spreadsheet. But course titles and course descriptions are not standardised and we are fairly much reliant on how course providers have chosen to describe their courses.
We hope to develop a better solution for accessing locally provided data on courses.
Better local labour market information
Obviously the ability to provide local labour market information is dependent on data sources. In some cases we have city wide data, in others regional. We would like to provide more up to date local data and once more are working with colleagues and organizations in our city partners to make it easy for them to provide such data.
Access to more videos
Our videos around different occupations are from the iCould charity. We hope to be able to provide a wider range of videos in the near future.
Save and return
We are working on functionality to allow CiCi to record prior interactions and remember these when a user returns to CiCi after a previous session.
Assessing your skills and possible new occupations
This will be a major development. We want to be able to predict the stability of existing employment, and then based on a skills assessment, suggest possible new occupations. This will be facilitated by the NESTA algorithm expected to be released at the end of January 2020.
Functionality to produce transcripts of CiCi interactions
We have had requests that users should be able to save a report on interactions which could be shared with a careers adviser or job coach, if the user wished for that. We are working on this functionality.
Coaching can be a powerful process in supporting users in planning their future jobs and careers. However, Artificial Intelligence and chatbots have limited abilities in this area. We are looking at how we can increase motivation and support through ‘nudging’ and access to micro units of learning materials.
We aim to increase the accessibility of CiCi through an optional audio interface in the future. We also aspire to developing multi-lingual versions of CICi
Newcastle, Derby and Bristol have supported us in developing the pilot and beta versions of CiCi and we hope to continue to work with them in the future. But we also wish to extend our reach and are in discussions with organisations from a number of further cities.