Approaching Microservices with Decisions and Standards

Rate this:
Total votes: 0

The mortgage industry – like most others looking to FinTech – have been exploring a variety of emerging technologies. All the usual suspects are included: artificial intelligence, machine learning, robotic process automation, blockchain, etc. Surprisingly, the exploration of APIs and microservices has been at the top of the list for many. Coming from a technical background I was somewhat surprised to hear these brought up. However, it became increasingly clear how important this concept truly is to the industry.

The drive for APIs was largely a result of the increased amount of data available to businesses and consumers. New and better ways have been required to access, filter, analyze and ultimately get value from the vast amount of data available. MuleSoft did a survey of 650 IT leaders and found that APIs and their integration possibilities were bringing real value to businesses in a variety of ways: integrating software-as-a-service (SaaS), making them more agile, and providing the opportunity to disrupt their business and create new revenue streams.

As I discussed in a previous white paper, RegTech is dependent upon utilizing the cloud to access and share data so that questionable transactions can be flagged in real time. Efforts are underway to standardize and enable B2B and B2C transactions to connect more players for this and other needs. So where do decisions fit in this space?

I believe decisions will play a key role in the evolution of APIs supporting microservices with the concept of decisions as a service. As Nathaniel Palmer at bpmNEXT 2018 said:

“Decision as a service for both commercial and government services will become a primary delivery model, allowing decisions (and the automation enabled by them) to be easily plugged into applications; this will require more of a business-first, model-driven approach rather than having decisions built in code by developers.”

We know already that decisions and their delineation with the DMN standard clearly support the business-first, model-driven approach. This will allow for APIs generally and microservices specifically to be created to serve a variety of current needs, enable new opportunities, and give an Enterprise the agility they all seek. DMN 1.2 has recognized this need and expanded the use of a Decision Service in the standard. It was possible in DMN 1.1 to specify an interface that allowed a decision model to be invoked by an external architecture like a Business Process Management System (BPMS). However, DMN 1.2 extends this so a decision service can be invoked withina decision model.  This means a decision service can be invoked from multiple places within the same Decision Requirement Diagrams (DRD) or even from a different decision model. This acknowledges the important realizations that:

  1. Decisions can be used in multiple places
  2. Decisions may be independent of the rest of a DRD
  3. Decisions and the accompanying models may be created and used by different parts of the Enterprise.

This corresponds nicely with the needs of microservices (that may be tied to a decision service within a model or the full model). For example, an API might allow a Loan Origination System (LOS) to connect to a platform to determine the pre-approval status for a new borrower. The full model determines the status. A decision (or micro) service in that model may retrieve and assess the borrowers FICO score. This microservice is one that would be used in numerous decision models. Externally, we have all the usual kinds of service deployment options available for the API (Open API, REST API, etc.). 

The most important factor here is something I’ve previously discussed. The Mortgage Banker’s Association standards group (MISMO) will soon be announcing DMN as the recommended standard for decision modeling, execution and exchange. The truly exciting revolution is the combination of a business standard – MISMO’s mortgage industry data standard – with a technical standard – OMG’s DMN. That means we can write decisions using the language of the mortgage industry. Decisions written in this fashion can be exposed via an API that is immediately useable by players in the mortgage industry! At DecisionX we have successfully deployed the early concepts of a Digital Decision Library for the mortgage industry than can take MISMO-compliant loans as input to render various decisions.

An approach like this has many advantages:

  1. It reaches more Enterprises. A pay-as-you-go Decision-as-a-Service (DaaS) model eliminates the need for costly platforms and often complex implementation.
  2. There is no more vendor lock-in. By using a decision service - when you need it - you can always remain current.
  3. It facilitates customization and personalization. These services can be tailored so they specifically meet the needs of a business as opposed to having to work within the constraints of a vendor platform.

While the idea of an API may not be emerging technology in and of itself, build microservices within them using standards and new approaches to technology certainly make this an exciting prospect for years to come.



Join the Discussion

Remind me later

Free training!

Want to sample our training?
Attend our Open House for immediate access to sample some of our newest courses. 

Schedule an appointment with a training advisor to learn more about our certificate programs.

Act now. The Open House is only available for a limited time.