It is not the strongest of the species that survives, nor the most intelligent. It is the one that is most adaptable to change.
Business policies are often hard coded into application. Externalization is the process to move them outside the program in the form of rules.
Externalization can be gradual, prioritizing volatile policies to accelerate payoff.
Tests shall be systematic to avoid errors.
The rule engine is responsible for executing the rules.
The engine supports different algorithms suited to different applications (tree, table, forward-chaining, backward-chaining, fuzzy rules, custom).
The engine smoothly runs as micro-service (or the application embeds it).
Users dispose of tools to compose and edit rules, simulate and debug execution, schedule activation and track changes.
Most rules are conveniently carried in plain spreadsheets. Complex cases demand a DSL (Domain specific Language).
Artificial Intelligence (AI) contributes further rules inferred from database and big-data stores, like, for example decision trees.
The rule engine automatically imports such rules to optimize decisions.
In addition to flexibility and adaptability, you will get tempestive forecasts by trying new policies and ruleson your actual and live data.
You will reduce the effort in application maintenance (typically consuming most IT budget) and to simplify communicate with IT team.
Systematic use of AI to extract rules results in continuity of optimized results.
Documentation is traditionally dated and incomplete, while rules, when used, constitute the live and actual documentation of business policies.
Can rule externalization be applied on shelf software? How shall I proceed if my application was produced by a third party?
In order to externalize rules you shall be in control of the application. This means not only dispose of source code but also being able to modify its structure. This is seldom the case with shelf software.
But we also fave few shelf products that are natively rule-enabled and can possibly match your needs. Please, take a look at that section.
Isn't rule externalization long and expensive? Can I afford it?
Cost and effort obviously depend on the the size and complexity of your application, extracting all rules in a single step will be a very massive and demanding project.
On the other side identifying the most volatile policies and proceeding incrementally is quite affordable and produces immediate return with little investment.
Are results guaranteed?
Each project undergoes a feasibility phase (free of charges) resulting in a multi-step plan. Each step externalizes a chunk of business logic.
Results are guaranteed: you has nothing to pay if you are not satisfied and prefer to stay rules embedded in code.
Rules based systems have been successfully used since 1980, but were limited to specific applications for two main reasons: computational resources and lack of business expertise.
What about new applications? Why external rules are not applied since inception?
This is surely the best alternative when possible.
If you are starting the development of a new application this is the best moment to introduce flexible rules.
Service Charging & Billing Solution
Service billing is usually complex: different rate plans, different charging policies (pay as you go vs. subscriptions) exceeding consumption, bundles, options, discounts, fidelity, floating rates and so on.
SolidBilling is a shelf solution for any kind of service billing. The rule engine transforms generic business events into charges, that are then periodically aggregated and billed.
CPQ (Configure, Price & Quote) Assistant
Product customization is a winning strategy to defeat competitors. It is today widely used even by B2C industry (automotive, insurance...). But configuring products and producing custom quotes demands knowledge and takes time. On the other side speed is essential in sales cycle.
Quotty produces accurate and highly configured sales quotes based on customer requirements and needs. It employs custom rules to determine optimal configuration and avoid conflicts. Quotty helps channels to sell faster, bigger and more.
An inference-driven chatbot
You have probably already used a chatbot. It is a computer program that mimics a human and interacts through a dialog. Chatbots are typically used for customer services or information acquisition but most of them are search-based, this is they look for a similar case for example in a problem database.
Binderbot on the other infer, from the set of defined rules, the correct sequence of questions to be asked to the user to reach the required goal through the best path.
