Get ready to Code, Innovate and Create at TEQnation online (May 10–12)

Event

Published in

ABN AMRO Developer Blog

5 min readApr 30, 2021

TEQnation is the biggest multitrack Developer Conference in the Netherlands. This year TEQnation is fully online (May 10–12). Three days of developer mayhem! Every day between 16:00 and 19:00 TEQnation will deliver quality content. At TEQnation you can broaden your knowledge at our hands-on labs, technical sessions, live demonstrations, and workshops, which are all focused on the technology of the next era. TEQnation 2021 will bring more than 1.000 developers and the best speakers from all over the world. Register for TEQnation here (for free!) and read more about the event below.

Interesting talks at TEQnation 2021

You can have a look at the full program of TEQnation, but we have highlighted two, in our opinion, must-see talks!

Transaction Monitoring 2.0 with MLOps

Lars Haringa and Jorg Klein of ABN AMRO will talk about data science relating to transaction monitoring (Room 1, Monday, May 10, 17:10). Read their abstract below:

Hardly a week goes by without news of money laundering affecting a financial institution. As gatekeepers of the financial system, banks play a crucial role in finding and reporting instances of financial crime. In this talk, we describe the challenges of detecting money laundering and outline why employing machine learning techniques is critically important to identify complex and ever-changing patterns.
We will present an overview of data science work related to transaction monitoring, and deep-dive into some of the challenges that we faced at ABN AMRO as well as the solutions we came up with. This includes how we continuously enhance an existing rule-based system with machine learning, as well as a method to deal with sudden data shifts while running in production. We will also discuss an in-house-developed model explainability technique that helps our transaction monitoring analysts focus their investigation and better understand what caused the alert.
A key capability for our enhanced transaction monitoring setup is how we facilitate both monitoring and retraining machine learning models live in production. MLOps has become the de facto way of working for teams that need to bring ML models to production. But how do you actually move from a few single models created by data scientists on an experimentation environment to dozens of models that are fully automatically trained, tested, deployed, monitored, and retrained — while ensuring traceability and reproducibility? We will illustrate the journey organizations go through as they move towards the next MLOps maturity level.

Personalization as a Service, a Data Journey (Keynote day 2)

Anne Feringa (ABN AMRO), Giuseppe Guida(ABN AMRO), and Wesley Delpeut (Capgemini) will talk about something new and cool: personalization as a Service (Room 1, Tuesday, May 11, 15:35). Check out their abstract below:

Marketing’: A dirty word, many would say, and how does that work at a bank? In ABN AMRO’s case, we use marketing automation to provide our customers with the best experience and the most suitable products and services. We want to optimize your menu based on your most used interactions, deep link you to the self-service page in case of need, and help you find the best suitable mortgage. But, how do we actually achieve all of that, and what is in the pipeline? It all starts with the data at the foundation of it, a smart choice of SaaS solutions and APIs to enhance and personalize experiences. In this talk, we are not going to talk about customer journeys but take a slightly different perspective and tag you along on the data journey.
Finding data in a big and regulated financial organization can be difficult, the data sources are innumerable and all of them could be useful to provide better services to our beloved customers. But how can we find the right data and how can we make good use of it?
ABN Amro has designed the perfect big data journey: 1 Marketplace, many Domain Datastores.
The journey starts as in any good recipe, in the right and sole Data Marketplace! A virtual repository of the banking data coming from the thousand’s applications used in the daily business. In a subscription fashioned manner is possible to simply reach and use what is needed for a Domain Data Store to provide insights. This is the first topic we will have a closer look at.
There are many Domain Data Stores spread across domains and departments that can then simply reshape, remodel and compute new and meaningful analytics for reaching our customers in the most efficient way! According to the Mash Data Architecture, each Domain Data Store is a mix of Big Data Technologies, Center of Design, and an Inner Source Development Center that gives meaning to petabytes of data.
The other component of the marketing automation chain that we’ll take a closer look at is the API layer that provides personalized content to the channels of ABN AMRO like the mobile app, internet banking, and call center. These APIs and underlying components process millions of requests a day and are developed using a cloud-native approach. This is all powered by the architecture that utilizes the serverless stack of AWS: APIGateway, Lambda, Aurora, and DynamoDb to name a few. The infrastructure is defined in code (Infrastructure as Code) and generic data structures enable content to be presented regardless of the channel it is presented on. During the development of these APIs, many challenges were faced and we’ll share the lessons learned of developing an enterprise-scale API layer.

Click here to look at the full program of TEQnation.

Register and join!

Don’t hesitate, register for TEQnation here (for free!).