Context

All organisations have policies, typically written out in meticulous detail in a number of different places. The policy is likely to be a combination of the organisation’s own policy and that of legislation that the organisation must adhere to. All organisations need to handle changes to their policies, whether caused by internal or external factors. There could be policy around the level of service that a customer is entitled to, or how a supplier is onboarded, or what benefits an employee is eligible for. Groups within an organisation will be experts in various policies. In its most basic form employees leaf through policy documents to find an appropriate response to a query – this might provide acceptable information, but it’s a slow process and is unlikely to provide consistency and transparency.


Rules Engines

The modern organisation has automated policy into a Business Rules Management System (BRMS), also referred to as a rules engine. Rules engine technologies have been around for a number of years known as Expert Systems in the 1980s, and Decision Management Systems in the 1990s. Current systems often include an element of Artificial Intelligence. A rules engine is there to assess available data – from a citizen/customer/supplier/employee, and other systems – and to reach conclusions and decisions. As rule engines technologies have evolved, organisations are reducing reliance on employees manually leafing through policy documents or information contained within their heads and are moving towards automated processes with intelligence embedded within them.


All good so far, so the problem has been solved?

Partly, yes, but not completely. Here are some questions to consider:

  • Can rules be created and updated easily?
  • Is there a single source of truth shared by different channels?
  • Are the decisions fully auditable?


Time to introduce Intelligent Advisor. Firstly, what is it?

Intelligent Advisor is a BRMS that allows for the modelling and deploying of rules. Rules work together to reach a top level goal which is also a conclusion, perhaps around eligibility for a benefit, a visa, or a discount. The rules are contained within a policy model (or rulebase) which can hold anything from a handful of rules to thousands of them. Here’s a very simple example:

In the example above, if both the conditions are true then the conclusion is proven to be true. Each condition can, in turn, be its own conclusion with a number of conditions underneath; so a hierarchy of rules gets built up that is, in theory, endless. In the example an and is used although you can also use an or, and you can nest rules like this:

What might surprise you is how clear the rules above are. They are authored in natural language using tools that many of us are familiar with, such as Microsoft Word and Excel. The rules are very readable to a non-technical audience, allowing the policy models to be owned by the business, specifically policy owners, subject matter experts, or business analysts. This means no coding and no liaison with the IT department which might be slow to respond and expensive to use. So, to answer the initial bullet, yes – rules can be created and updated easily.


Single source of truth. 

Rules can be centralised and accessed through multiple channels. By channel we could mean a customer on the phone, or a citizen accessing a webform, or a chatbot, as well as others. Broadly speaking there are two ways to access a policy model:

  1. Through a user interface (UI) that is available in external websites and internal systems. 

Example: an employee answers questions on the intranet to find out eligibility for a workplace benefit.

  1. Via web services calls where other systems access the policy model with a payload to get a response. 

Example: a Customer Relationship Management (CRM) system calls the policy model to determine a customer’s discount on a new product.

The point being, the same rules are being accessed – regardless of channel – and will therefore provide a consistent, transparent and repeatable response. Intelligent Advisor provides a platform for creating the rules as well as the creation of a UI. The default UI is very flexible and can be branded to match corporate styles.


Auditability

What happens if a decision is queried? Do you have all the inputs to find out how the decision was made? By default, Intelligent Advisor will create what’s known as an explanation of how its goals were reached. These explanations can be saved and made available to auditors or super users to interrogate.


Intelligent Advisor sounds very much like Oracle Policy Automation or OPA.

You’re one step ahead! The name was changed right at the end of 2019. Most of us have adapted to the new name, but we still do use the OPA name from time to time. The solution itself is provided by Oracle and is typically deployed to the Oracle Cloud.


Conclusion

This blog has given you a brief overview of policy/rules and how organisations have moved away from paper and into automation. Questions have been asked about the automation tools used and how well they have solved the common problems that persist: 

  1. Can rules be written and updated easily?
  2. Is there a single source of truth?
  3. Are all your decisions auditable? 

We’ve introduced you to Intelligent Advisor, a modern rules engine that addresses these problems.

If you’d like more information then we run Intelligent Advisor webinars. They are listed here: https://www.eventbrite.co.uk/o/magia-consulting-23369950486

In addition we are more than happy to take queries: jonathan.watson@magiaconsulting.com


About Magia Consulting

Magia Consulting is a technology solutions provider and consultancy that specialises in intelligent automation technologies to empower organisations to convert complex business requirements and policies into simple to use, expert digital solutions.