By **Bhupendra Parmar**, Public Sector Sales Executive
By Bhupendra Parmar, Public Sector Sales Executive
Far-fetched claims suggest AI could learn to recognise galaxies, while more well-known possibilities propose driverless AI controlled cars and AI-based technologies diagnosing terminal diseases. However, at its most basic, AI refers to a computer system’s capacity to perform tasks typically requiring human intelligence.
Over the last few years, the UK government have made artificial intelligence a significant policy initiative detailing how artificial intelligence and machine learning (ML) can automate defined, repeatable tasks, augment decision-making and enhance understanding in policymaking and public service delivery.
However, you can’t just deploy AI/ML technology and expect it’s going to make everything better. You need three things to make it work:
The availability of large data sets can help train AI systems to recognise patterns in those data sets with little human direction, leading to better, more informed decisions over time.
AI does not come ready to work out of the box. In most cases, you need to integrate the right data sets, apply the appropriate topic modeling and choose the correct use cases.
That’s what we’re working on at 8x8. We have massive data sets across our combined Unified Communications and Contact Centre platform, spanning voice, video, messaging and contact centre information. We have pooled this data together and aggregated it, made it useful to glean insights and improve customer interactions.
We have trained our virtual agent to start customer conversations by predicting why they are calling so it can get right to the point on its first try using the data at hand.
We’ve partnered with Google Contact Centre AI to understand the voice transcription, the topic modeling and the caller profile. We use their dialogue flow and set up an interaction with the knowledge-base, and present the right information back to the customer.
Some customers won't be fully satisfied with a response from the virtual assistant, or they may call back again. When that happens, we use the customer’s knowledge graph to understand that they’ve called previously and presumably their prior service visit was not successful. The customer can be directed to a knowledgebase article, or the call is escalated to a real service agent, and we use AI to do just that.
The call is split through the transcription and interface to the Google Contact Centre AI backend, and we present the knowledge-base article to the agent, to help provide a quick response to the caller. If the agent cannot resolve the issue, then an expert within the company can be located to help solve the customer query.
This is where it is essential to have a single communications platform with unified communications and contact centre capabilities.
Agents can also locate information from public chat rooms, which have a wealth of knowledge about a particular issue or topic. The same AI topic modeling is applied, allowing the AI to query the rooms and make suggestions regarding which expert to call, if a subject matter expert outside the contact center is needed.
Local councils can take this approach to start deploying AI-based virtual assistants within the contact centre, combining data from across the organisation to help obtain the right answers in real time.
8x8 provides this highly scalable, consistent data platform, that processes real-time and historical data, effectively leverages the compute power of the public cloud and provides immediate responses. It’s not what you get in a typical chatbot solution, with generic answers based on natural language processing.
We’re dedicated to helping our public sector clients unlock the positive potential of Contact Centre AI.
Request a demo Our dedicated public sector team can be contacted at uk-publicsector@8x8.com
For more information call 0333 043 8888 or visit 8x8.com/uk-public-sector
