by Brendan van Staaden
Brendan is an interaction automation and customer experience expert and Managing Executive of MoData Interactive
During a recent workshop, I was asked about the key elements of decision support that combine to influence decision intelligence and how this affects conversational business applications (CBAs) and the management of the enterprise content.
When it comes to a sound decision support system, it is always about the data! Whether the data source is internal, external, structured, or unstructured, a comprehensive and current data set is fundamental to making informed decisions. Modelling this data to analyse and predict outcomes is only possible once the desired outcome is adequately defined and the nature of the data understood in terms of how it will influence the outcomes once applied.
Algorithms used to automate tasks and make decisions can be used to sort the data, identify behavioural and usage patterns, and make predictions. As powerful as these algorithms are, they are only ever as good as the data set, they are trained and reliant on. At the enterprise layer, it is crucial to manage content in the most effective way, to render the data meaningful for effective algorithms and sound modelling. Gone are the days where enterprise content management was done solely for the sake of compliance and data security. As much as these are still valid, the ECM regime extends to include effective data structures that will inform automation strategies and decision intelligence.
User interfaces and expertise
User interfaces and expertise are two underestimated aspects of an effective decision support protocol. When the interface is easy to understand and manipulate, the users can explore data models in meaningful, interesting, and accurate ways. The better the UI, the more inclined users will be to experiment and play to extract the most telling outcomes. Having the right expertise to understand and interpret the data models and effects, is crucial to the successful decision support system. Users need not be super scientists, but they will require specific skill sets when it comes to understanding the data, the application, models, and algorithms. I have worked with some of the most interesting folk in these spaces where making the data dance is part of the game.
All these elements of the decision support system combine to deliver decision intelligence, the simple ability to make better decisions using the data, the models, algorithms, and expertise, specifically aimed at helping businesses to improve performance, reduce costs, make more informed business and strategic decisions and to best utilise and deploy existing resources without having to reinvent the wheel.
Conversational business applications
So how do conversational business applications support decision intelligence? Well, firstly, CBA’s make it easier for users to access and interact with the data sets we speak of here. By using natural language processing (NLP), natural language understanding (NLU) and other related AI techniques, CBA’s allow users to ask questions about data in a natural way, without having to learn complex query languages. This can make it easier for operators to find the information they need to make informed decisions. Of course, the success of this again remains critical in relation to how data is managed, structured, and stored. A giant step in the ability to query data sources to support decision intelligence.
Secondly, CBA’s may help users to better understand and interpret the data sets through the delivery of advanced visualisation and interactive data manipulation toolsets. Through the provision of these advanced applications, CBA’s can better render, and help users see patterns and trends in data that they might not be able to observe otherwise, giving them a better “sense” of the data to make more informed decisions.
One of the key elements of conversational business applications, is the automation capability which goes beyond rule-based robotics to understanding and conversational, contextual interaction, which will help users to automate decision-making processes. By using machine learning and other AI techniques, CBA’s will allow learnings from past decisions and make recommendations for future decisions. This will free up users’ time for them to focus on other tasks, and it will invariably help improve the accuracy of decisions.
Working in tandem with generative AI
Conversational business applications working in tandem with generative AI capabilities and toolsets such as ChatGPT and the like, allow LLM approaches to enterprise data, making it easier for internal operations to improve their knowledge base and respond to customer queries more accurately, efficiently, and consistently. For example, a customer service representative can use a CBA to answer customer questions about products or services. The CBA will access data from the company’s CRM system to provide the representative with the information they need to answer the customer’s questions.
Sales teams, and sales management may use CBA’s to identify potential customers who are likely to be interested in the company’s products or services. The CBA can access data from the company’s marketing automation system to identify these potential customers, related propensities, and strategic intent. Yet again, the dependency will always remain on the data that the companies keep on their customers in their requisite ERP and CRM back-end applications. A typical risk management scenario unfolds where a risk manager can use a CBA to assess the risk of a particular investment. The CBA will access data from the company’s financial systems to assess the risk of the investment and provide variable outcomes dependent on the nature of the enquiry.
Decision support in creating scenarios based on intelligence collected and collated across enterprise-wide data outputs will always inform sensible approaches to any business decision where a variety of unknown factors aren’t always taken into consideration when making complex judgements. In today’s technological world where a plethora of new and legacy systems combine to try and deliver sensible customer outputs, traditional thinking may not be enough to inform a modern and progressive strategy when it comes to improving customer engagement and experience.
Conversational Business Applications, applied to well designed and properly managed data sets will invariably allow a sound decision support outcome through decision intelligence from mining logically structured content
