With AI constantly changing, how difficult is it to ensure that AI use continues to be relevant to customer needs?
When we talk about AI in business, it’s almost always associated with machine learning (ML), which are computer algorithms designed to carry out one task and become increasingly good at it as they repeat it again and again.
Some simple examples in the context of CX might include:
- An ML algorithm to understand a customer complaint and direct them to appropriate help, reducing the time they need to spend waiting for help (a chatbot).
- An ML algorithm to recommend the products and services a customer is most likely to find useful, reducing the time they need to spend searching on a website (a recommendation engine).
- An augmented reality (AR) application that lets you see how products from a retailer might look in your own home before you buy them or that lets you try on clothes in a “virtual dressing room.”
These are all simple methods of using AI to create more streamlined and rewarding customer experiences that are used by thousands of businesses around the world. In a nutshell, AI and ML should be continuously being refined to be relevant and effective in solving current and evolving challenges, including:
- Being in-step in the way the world communicates. Billions of consumers are moving from traditional channels like email and phone to modern channels like social, messaging, chat, and text.
- The amount of experience data (likes, comments, ratings) customers share across digital channels is unprecedented and unstructured. With siloed teams and disconnected technology, companies struggle to manage this data.
- Effectively addressing changing consumer expectations. Consumers value instant, personalised, proactive, and consistent experiences above all else.
Brands looking to increase revenue, reduce costs, mitigate risk, and deliver memorable customer experiences, need to invest in solutions that help eliminate disconnected processes, siloed teams, and disparate technology with one unified, AI-powered platform.
What is the difference between bound vs. unbound data and why is that important when training AI?
Bound data is finite and unchanging data, where everything is known about the set of data. Typically, Bound data has a known ending point and is relatively fixed. An example is what was “last year’s sales numbers for xx car model.” Since we are looking into the past, we have a perfect timebox with a fixed number of results (number of sales).
Traditionally we have analyzed data as Bound data sets looking back into the past. This uses historic data sets to look for patterns or correlation that can be studied to improve future results. The timeline on these future results could be measured in months or years.
For example, testing a marketing campaign for the XX car model Y would take place over a quarter. At the end of the quarter sales and marketing metrics are measured deeming a success or failure for the campaign. Tweaks for the campaign are implemented for next quarter and the waiting cycle continues.
The question is: Why not tweak and measure the campaign from the first onset? Our architectures and systems were built to handle data in this fashion because we didn’t have the ability to analyze data in real-time. Now with the lower cost for CPU and explosion in Open Source Software for analysing data, future results can be measured in days, hours, minutes, and seconds.
On the other hand, unbound data is unpredictable, infinite, and not always sequential. The data creation is a never-ending cycle. For example, data generated on a Web Scale Enterprise Network is Unbound. The network traffic messages and logs are constantly being generated, external traffic can scale-up creating more messages, remote systems with latency could report non-sequential logs, and etc. In other words, all the data points are unpredictable and infinite.
This all means that we need better systems and architectures for analysing Unbound data, but we also need to support those Bound data sets in the same system.
Data is the lifeblood of modern artificial intelligence. Getting the right data is both the most important and the most challenging part of building powerful AI. Collecting quality data from the real world is complicated, expensive and time-consuming.
Often we see workforce resistance towards deploying AI. How differently should customer service agents should be trained to augment their jobs with AI?
For a lot of businesses, the purpose of AI in customer service is to automate easy, boring tasks. This aims to ease the load of real agents, enabling them to focus more on complicated tasks that AI can’t handle.
With repetitive cases being automated, brands need to keep motivation and energy levels high by supporting their staff with adequate training that will prepare them for bigger challenges.
In addition, there’s also the challenge of working with AI to provide great customer experience. AI should be able to point to the best courses of action and product recommendations, while humans should be able to make the right calls. When customer experience agents are equipped with the know-how to seamlessly work with AI, resolution times will drop, employee burnout will be avoided and everybody wins.
It is also important to highlight that as customers increasingly turn to self-service for simple issues, they will primarily rely on agents for more complex inquiries. By using AI to surface relevant knowledge and customer data in real-time, agents will be more prepared for these complex issues – and also able to devote more focus to connecting with customers through warmth and empathy.
AI-enabled customer experience solutions also help drive a favourable impact on the employee experience; the effort today’s agents spend fumbling through knowledge bases and CRM systems tends to be one of the greatest drivers of dissatisfaction and churn.
What guidelines/regulations that should be in place for AI to flourish in the future?
In the global discourse on AI ethics and governance, Singapore believes that its balanced approach can facilitate innovation, safeguard consumer interests, and serve as a common global reference point. Hence, Singapore is not rushing to set AI regulation.
Singapore believes that even as it develops its digital economy, a trusted ecosystem is key – one where organisations can benefit from tech innovations while consumers have the confidence and trust to adopt and use AI.
To advance the use of AI in the country, Singapore had developed AI Verify, an AI governance testing framework and a software toolkit. The former consists of 11 AI ethics principles* which are consistent with internationally recognised AI frameworks such as those from EU, OECD and Singapore’s Model AI Governance Framework. Through AI Verify, organisations can validate the performance of their AI systems against these principles through standardised tests.
To shape the future of AI standards, IMDA has also recently set up the AI Verify Foundation, which drives collaboration with the global open source community to develop AI Verify testing tools for the responsible use of AI. The Foundation will boost AI testing capabilities and assurance to meet the needs of organisations and regulators globally.
*The 11 AI governance principles include transparency, explainability, repeatability/reproducibility, safety, security, robustness, fairness, data governance, accountability, human agency and oversight, inclusive growth, societal and human well-being.
