This sample application shows one of the ways IRIS can be used to create a service that:
The following image shows the architecture of the solution:
You can see a video of this demo on YouTube.
To expose this new service, IRIS still relies on other systems such as the core banking system and the CRM. To interoperate with these systems, IRIS uses business process orchestration, business rules and look up tables (for coding system normalization).
When the service is operating, clean, normalized data starts to flow through IRIS. Instead of throwing this data away, IRIS can easily store it on a normalized data lake. This data can be used to monitor the business in real time, monitor the ML model performance over time and also to train better ML models.
There is no need to do the ETL (Extract, Transform and Load) all over again. Clean data is the side effect of using IRIS to expose your service!
The application brings a POS (point of sale) simulator. It is a simple Angular UI that we can use to swipe our cards and simulate transactions.
To just run the application on your PC, make sure you have git and Docker installed on your machine.
You will need to fork or clone this repository to you local machine to get the entire source code. So, go to your git folder and run the following:
git clone https://github.com/intersystems-community/irisdemo-demo-fraudprevention cd irisdemo-demo-fraudprevention docker-compose up
That should trigger the download of the images that compose this application and it will soon start all the containers. When starting, it is going to show you lots of messages from all the containers that are staring. That is fine. Don't worry.
When it is done, it will just hang there, without returning control to you. That is fine too. Just leave this window open. If you CTRL+C on this window, docker compose will stop all the containers (and stop the application!).
After all the containers have started, open the application landing page on http://localhost:9092/csp/appint/demo.csp.
Use the username SuperUser and the password sys. This is just a demo application that is running on your machine, so we are using a default password. The landing page has instructions about how to use the demo application.
You don't need all the source code to run the application. But the application relies on a folder structure that allows:
It is just easier to clone the repository and get this folder structure "out of the box" instead of having to recreate it.
Here are some additional resources:
There are other IRIS demo applications that touch different subjects such as NLP, ML, Integration with AWS services, Twitter services, performance benchmarks etc. Here are some of them:
Please, report any issues on the Issues section.