Service Organisations as Intelligent Agents

The digitisation of service organisations has progressed to an advanced level. Bi-directional “customer < – > organisation” communication, service management and transacting can now occur via multiple digital channels.

What is the focus of service organisations now? Winning the race to make this digital capability pay as service organisations transition to being automated Intelligent Agents. This will take a few years still as AI improves and ever more data sources and value chain partners fully digitise and offer open platform capability, but the journey is well underway. Learning Machines has been at the forefront of enabling our customers to make this transition. What is it all about?

Paraphrasing Wikipedia:

“In artificial intelligence, an intelligent agent (IA) refers to an autonomous entity which acts upon an environment using observation through sensors and consequent actuators, directing its activity towards achieving goals. Intelligent agents may also learn or use knowledge to achieve their goals.”

Another way to represent this is as the process of:

An Intelligent Agent senses what is happening in its environment, it has a model of the world and can make predictions about what may occur next and what effect an action it could take would have upon its environment and its state within that environment, all toward achieving the agent’s goals.

So how does this play out in service organisations? A more suitable process depiction is:

Event

Collecting all available information and making it immediately available to intelligent downstream processes: transactions, service usage, browsing of your website, app usage, phone calls, emails, chatbot engagements, sms’s, 3rd party data such as credit records, relevant social media interactions etc.

The more events you make available the greater your organisation’s ability to react to what is happening in the world.

Enrich

Enriching the event information with everything you know about the customer, with the information already being available and/or calculated in real time based on information within the event: current product usage and usage patterns, sentiment of recent interactions, in-flight processes, credit history, account status and predictions such as next best action, propensity to purchase any of your products etc.

Decide

Combining all information with an integrated view of customer and organisational goals to decide upon the next best action/s: send a message to the customer, offer products, suggesting changed behaviour, generating leads, initiating collections processes etc. whilst taking into account rules to manage message/marketing fatigue, privacy etc.

Act

Making it simple to execute any one of 100’s of actions toward achieving goals across channels, products, services and processes.

The more actions you make available the greater your organisation’s ability to act upon the world toward achieving goals.

Learn

Analysing whether actions lead to the expected changes in the environment such as the customer purchasing a product or not, changing behaviour or not etc. and creating training data for algorithms to learn how to better predict outcomes from actions based on a comprehensive understanding of the customer and organisation’s current state.

Achieving the above is not easy, not by a long shot. The Learning Machines team has been delivering highly complex solutions for decades, but in all honesty creating the Intelligent Agent Service Organisation is a next level of complexity.

It involves collecting millions of events from a broad range of sources and making all available on publish/subscribe systems such as Kafka, calculating and making available 1000’s of facts and predictions about the world with sub-second latency, deploying and managing hundreds of machine learning models making predictions, 1000’s of facts on millions of customers over many years to train predictive models, enterprise integrations to easily initiate 100’s of business processes, managing and updating business rules to enable decision making etc. etc.

This is why the Learning Machines team consists of software engineers, data engineers, data scientists, architects, cloud engineers and project leads that have decades of experience in delivering complex enterprise solutions.

After many months of hard work in collaboration with customer teams we are starting to bring to life the Service Organisation as Intelligent Agent in South Africa.

Our client’s automated systems have now made millions of decisions based on what is happening in their world right now and taken millions of actions toward achieving customer and organisational goals. Now they can use the resulting data to know more, predict better and beat the competition in achieving customer satisfaction.

How far are you along this journey? Want to move faster? Contact Learning Machines today. We exist to help automate production, distribution and service in South Africa and beyond.