CiCi: Fulfilling the Promise of Career Chatbots

The world has changed dramatically in the last year. Societies everywhere have had to adapt and find new ways to prosper during the pandemic. Video conferencing, Zoom calls, Facebook and more regular use of innovative social platforms for online support and new working practices are considered the ‘new norm’. Technology is becoming even more ubiquitous. Digital growth, increased automation and artificial intelligence (AI) require individuals to be committed lifelong learners (ETF, 2020) [i] – thinking about their transferable skills, switching from at-risk sectors to remain in employment and/or simply browsing for new career ideas.

We live in an age of conversations often powered by artificial intelligence (AI) and machine learning. This may take the form of scripted customer service chatbots embedded in online banking or insurance systems, to non-scripted chatbots that help inform and support individuals’ health, well-being and other specific needs. New conversational interfaces are becoming part of our everyday lives. For careers support services, understanding how best to use open datasets in this context, capturing new and evolving labour market trends and new forms of career trajectories, is vital.

The impact of Covid-19 has resulted in rising levels of inequality with lowest earners more likely to have been affected by the pandemic than those on the highest salaries (Joyce & Xu, 2020)[ii]. There is a growing need to increase efforts in promoting learning and work opportunities. Individuals need easy access to trustworthy careers information, advice and guidance (Attwell & Hughes, 2019)[iii], particularly in volatile, uncertain and complex labour markets.

Digital technologies are deeply intertwined with human activities. We believe humans and bots can potentially work well together (Arthur, 2021)[iv]. New technology creates online spaces and places for constant interaction between humans and non-humans. The use of big open datasets has significant potential, including use of national and local labour market trends, salaries and job prospects, though has proved challenging in many countries (Řihová, 2016)[v]. This is further accentuated by the shock of the pandemic in key sectors. An over-reliance on traditional forecasting methods using historical data only is no longer sufficient. Instead, innovative approaches and transparent quality standards in the usage of AI and LMI data are now required. This will help rebuild knowledge and understanding of rapidly changing labour markets and the evolving opportunity structure in the years ahead.

A well designed chatbot system can respond to repetitive and typical questions and can improve the efficiency of a careers or employability service in freeing up time for practitioners to focus on more in-depth support for their clients. In England, NESTA (2020) – an innovation foundation – and the Department of Education (DfE) set a national ‘CareerTech Challenge’ through a prize competition on innovative uses or sources of LMI to improve careers information, advice and/or guidance.

CiCi[vi] – the powerhouse that supports your career™’ – one of twenty national finalists – is co-designed alongside professionally trained careers and employability advisers.[1] This theoretically informed prototype chatbot, currently being tried out in practice, combines AI and NLP to translate bite-sized careers information and advice to adults in three major English cities accessible throughout the day and night. The chatbot can be accessed on computers and mobile devices and can be embedded into web portals and social media. Designed to be flexible and agile, the software behind CiCi is modular allowing for rapid adaption and updating.

The ongoing development is testing the boundaries between online careers information and advice and professional career guidance, delivered by qualified careers and employability specialists. In particular, can the bot be trained using AI to learn its own limitations and to know when to refer the user to a careers’ professional?  With permission granted by the client, can the chatbot share information in advance of a careers interview, so that the person does not have to repeat what they have done already in their career search? CiCi provides reassurance to the individual user of professional support available locally, if needed. The chatbot also enables practitioners to have an added tool to work effectively with their clients and concentrate their specialist skills on those most in need.

It is time to reflect on the role digital technology should have in the future of education, training, employment and career guidance conversations, not only as a solution that enables the continuance of services during such a pandemic, but also for greater 24:7 personalisation of self-learning and reflection. Professionalisation of the workforce must include finding ways of co-designing and using chatbots and adding this to the practitioners’ expert repertoire of support services. Career chatbots hold great promise for organisations, professionals and their clients/customers. Now is an exciting time for practitioners, programmers and organisations to engage and stay ahead of the innovation curve.

Authors of this article Graham Attwell and Deidre Hughes are members of the innovation team who have developed CiCi. If you would like to find out more about their work, be sure to join OEB21. Meanwhile, to try out a working demonstration of CiCi, please visit: https://careerchat.uk/.

[i] ETF (2020). Innovation in career guidance. International trends and case studies in the European Union, internationally, and in selected partner countries. European Training Foundation. https://www.etf.europa.eu/sites/default/files/202011/innovation_in_career_guidance_vol._1.pdf

[ii] Joyce, R. & Xu, X. (2020). Sector shutdowns during the coronavirus crisis: which workers are most exposed, London: Institute for Fiscal Studies, April 2020 – https://www.ifs.org.uk/publications/14791

[iii] Attwell, G., & Hughes, D. (2019). Aprendizaje sobre carreras: datos abiertos e inteligencia del mercado laboral. RIED. Revista Iberoamericana de Educación a Distancia, 22(1). doi:10.5944/ried.22.1.22289

[iv] Arthur, M. (2021). Building a Bot to Beat the Bots: An Award Winning Approach from Across the Pond. Forbes.com, March 2021 – https://www.forbes.com/sites/michaelbarthur/2021/03/29/building-a-bot-to-beat-the-bots-an-award-winning-approach-from-across-the-pond/?sh=6d8a8e252a6a

[v] Řihová, H. (2016). Using Labour Market Information: Guide to anticipating and matching skills and jobs, Vol. 1, European Training Foundation, Cedefop and ILO – https://www.etf.europa.eu