A Human Resources robot: the virtual assistant

They might not yet be well-known as IPAs (intelligent personal assistants) but these virtual helpers have already taken over the world of internet-connected consumers. Any user of Siri (Apple), Google Now, Cortana (Microsoft) and Alexa (Amazon) has a virtual assistant at their disposal. By simply clicking in your browser or pressing lightly upon your smartphone, you can ask it a question. They will give an answer tailored just to you… or send a text message for you, or even determine the route you should take. Soon, they will be predicting your requirements by analysing your habits, reading your diary, assessing your profile, and picking up your past interactions progressively via machine learning. Currently serving as assistants-on-demand, they are set to become proactive, artificially intelligent conversationalists, capable of suggesting replies to send in discussions with friends, for example.

Why not in Human Resources?

These tools have hitherto been used mainly for personal purposes, but are now becoming widespread in the workplace too… so why not in Human Resources? Around ten years ago, we saw the emergence of applications combining a ticketing system and a search engine linked to a knowledge base. They sought to better handle HR requests in the context of an HR shared-service centre. The employee asking a question online would be encouraged to firstly find the answer themself in a list of FAQs before putting their query to a real-life assistant

Today, a virtual assistant can go further in automated support, via new technology and artificial intelligence. Most often, they appear as a 3D avatar representing a contact person who seeks to ensure that the user’s experience remains smooth and helpful. For example, if the assistant notices that no actions have been performed over a given period of time, they might kindly pop up to ask, for example, “Hello Chloé, what can I do for you?”. What can then follow is a user-assistant dialogue, made especially straightforward and intuitive by use of natural language and contextual analysis.

Capable of understanding users’ words and expressions, the assistant instantly analyses questions and swiftly answers them. Making use of their multilingual understanding and spell-checking faculty, they scour their knowledge base to formulate the most appropriate response. They can converse orally or textually and, if need be, involve images, sounds and videos thanks to their assimilation of rich media.

Anywhere, any time, any device

Virtual assistants can be used on any device (PC, smartphone, tablet, large screen, interactive terminal) and are available ATAWAD (anywhere, any time, any device) to support mobile employees. They make employee experience interactive, smooth and user-friendly, thereby ensuring that the system is embraced and that the return on investment is quickly profitable.

A robust back office is firstly required for this assistance to run smoothly and remain relevant. To ensure that answers are always helpful, there needs to be, crucially, a coherent knowledge base, involvement of other support channels, and high performance management tools.

Don’t be afraid…. The virtual assistant is auto-learning 

Via the knowledge base, the virtual assistant finds the right answers itself. However, these knowledge bases, which are often topic-based, vitally need to be as rich as possible. Rolling out the system can therefore be slowed down by this key requirement. This makes it important to set up and manage a process whereby the knowledge base is continuously updated. In the launch phase, the virtual assistant will frequently look for human help from experts. The answers they receive will then be tracked and saved to enrich the knowledge base. Furthermore, these knowledge bases can be built up in network mode, making use of SaaS solutions. By adopting a community-based approach involving several anonymous client bases, a common foundation can be shared, irrespective of company. The system is enriched as and when it is used through collaborative and collective intelligence.

To ensure that employees embrace the system and that their user experience remains smooth and useful, these virtual assistants can interact with other support channels. Virtual dialogue can thereby become an uninterrupted online chat between people. The virtual assistant’s back office can handle the flow of requests calling for varying levels of support, and send out tickets for deferred responses to be followed up later.

Lastly, the system can be monitored, analysed and assessed via management tools. This also helps ensure that users remain satisfied. Measuring the volume of requests by topic, measuring the processing time taken from when a query is submitted to when it is satisfactorily answered, listening to clients’ opinions via their appraisals and suggestions… these are some of the means to optimise the continuous improvement cycle of the virtual assistant.

 In conclusion

It is worth noting that the avatar’s simplicity can actually prevent the user from grasping the virtual assistant’s true power, complexity, capacity and relevance. Nevertheless, these solutions combine artificial intelligence to better understand users’ many natural languages, knowledge bases built collaboratively by both the virtual assistant and its human contributors, and management tools helping to ensure that users embrace the system and remain satisfied.

However, in terms of return on investment, how can its efficiency be measured?  Companies tempted by virtual assistants are bound to be asking themselves this question!