The Monthly Briefing

Back to basics

Personalised messages in BBVA app. The message shown in the image indicates to Ana that she has received a higher than usual income.

The basics are that your bank notifies you of important things. It is important to know that you have received a higher expenditure than expected or that you are at risk of being overdrawn, for instance.

At BBVA AI Factory we have been working on the data analytics behind this new feature integrated in the BBVA app. We are also very proud that a product based on Artificial Intelligence is accesible from Global Position, the first screen of the app. And not only that: we also have a TV spot! BBVA's latest campaign has focused on this new feature, which is already included in the new version of the app for all individual customers in Spain.

This is undoubtedly a firm commitment by BBVA to increasingly integrate insights based on data analytics. This new feature helps our customers to have better Financial Health and to know about important changes in their accounts as they occur -or even before they occur-. To do this, we have listened to our customers to find out which messages are most important to them, and we have also included others that are relevant because of the impact they have on Financial Health. One of the most remarkable aspects of this new functionality is that it generates messages based on the specific financial situation of each of our customers, so the messages are very personalised. And we send them when they really matter.

Every message is based on a different data engines: some are simply statistical and others are based, for instance, on balance prediction models. And then we have another data model that is responsible for prioritising these messages, using customer navigation data.

Congratulations to the Financial Health team at BBVA AI Factory, as well as the Proactive Advice & Personalization team and Client Solutions at BBVA, who have led this project.

"Three keys to working effectively on Artificial Intelligence projects"

On many occasions, the greatest impediments to creating Artificial Intelligence solutions do not lie in the capacity of highly qualified teams, but in establishing an effective way of working between the different professional profiles involved in the life cycle of analytical models. This is one of the main tasks we are currently tackling at BBVA AI Factory. It is a task guided by three concepts: simplify, accelerate and reuse.

My first direct contact with the AI Factory was in April 2020, in the middle of lockdown. I found myself with a team of data scientists who were extremely competent in creating AI models, but who needed to continue to push for common working guidelines in order to deal with the complexity - both organisational and technical - that exists in the Engineering domain. At the end of the day, models and data engines have to be integrated into the various channels of the bank to be made available to our clients. This makes it essential to work together as one team. The best models are those that reach the end user.

In this article, I delve more deeply into the three key concepts I mentioned earlier - simplify, accelerate and reuse -. With these ideas in mind, as COO (Chief Operating Officer) I join this great AI Factory team, recently set up as part of the BBVA Group, acting as a bridge between the worlds of Data and Engineering.

Ultimately, our purpose is none other than to squeeze all the potential offered by Artificial Intelligence, whether it be in decision-making, process optimisation or in the creation of products that offer added value to our clients. We are convinced that AI will bring significant benefits to society in general and to our clients in particular, supporting them in making the best financial decisions.

Check out the article published by Marta Sanz on Linkedin (in spanish).

Towards a better understanding of neural networks

One of our top priorities at the AI Factory is to keep up to date with the work being done in Data Science by colleagues from other companies and institutions. For this reason, we periodically hold internal meetings in which we learn from other experts in our field. A few weeks ago, Pablo Morala, Research Assistant in the UC3M-Santander Big Data Institute (IBiDat), presented us the idea of this recent article published on arXiv by the uc3m-Santander Big Data Institute.

This work explores how to build a mathematical framework relating neural networks and polynomial regression. Despite being one of the most widely used machine learning tools, neural networks are still considered black boxes due to their lack of interpretability, present difficulties when choosing their hyperparameters and structure, or the uncertainty in their predictions can be hard to model. With this framework, the polynomial regression coefficients can be interpreted in an easier way than neural network weights and the well studied statistical properties of polynomial regression can help the way in which we understand and use neural networks.

Further reading

+Experts analyse the EU's new Artificial Intelligence Regulation (OdiseIA)
OdiseIA (Observatory on the Ethical and Social Impact of Artificial Intelligence) organised a discussion panel on 26 April to analyse the recently published EU regulation on AI. Leading experts in the world of Data and Artificial Intelligence, such as Richard Benjamins (Telefonica), Lorenzo Cotino (University of Valencia) or our colleague Juan Murillo (BBVA), among others, focused on relevant aspects such as the risk levels of AI solutions that this regulation sets or the adaptation work that it will entail for companies and public bodies.

One of the most interesting aspects of this regulation is that it focuses on the risks and not on the technology itself. In other words, the level of risk depends on each specific use case. On the other hand, this new regulation aims to provide a legal solution to the gaps in existing regulation, and thus cover the various uses of Artificial Intelligence. The event was really useful to understand the general lines of this regulation.

+Beethoven's metronome (Almudena Martin-Castro and Iñaki Ucar)
For most of the history of classical music, the tempo -or speed of the music- was not unambiguously specified, but words such as Lento, Adagio, Allegro, Presto, etc. were used.In 1815 Maelzel invented the metronome, which put an end to this subjectivity for good. Beethoven not only implemented its notation in his next works but also marked it in his 8 previous symphonies. These time signatures are now considered by many to represent a very high speed, which would even make them sound unmusical, and no one plays it exactly as Beethoven specified. In fact, it is believed that Beethoven's metronome - one of the earliest in history - may have been broken.

Almudena Martín-Castro and Iñaki Ucar have published an article in which they model with high precision the functioning of a metronome and test different ways in which it could have been broken to give rise to these marks. The result of their analysis is so surprising that we can't tell you more, just keep in mind that: it includes data, technology and music... You have to read the article! You can also find out more about the project and the conclusions in this video by Jaime Altozano. Thanks for sharing Amanda Garci!


of the


"Without statistics, you’re stuck unable to know whether the opinion you just formed holds water. Without analytics, you’re flying blind with little opportunity to tame your unknown unknowns."

Cassie Kozyrkov, Head of Decision Intelligence at Google

In the article entitled How to spot a data charlatan, Cassie Kozyrkov identifies and differentiates the role of analysts -that help you come up with good questions (hypothesis generation)- and statisticians -that help you get good answers (hypothesis testing)-, and shares some tips for spotting data charlatans, who are neither analysts nor statisticians. As Cassie says, "while analysts offer you open-minded inspiration, statisticians offer you rigorous testing. And charlatans offer you twisted hindsight that pretends to be analytics plus statistics".

We love books

Via Getty Images

On the occasion of the World Book Day, held last 23rd April, at BBVA AI Factory we wanted to recover all those readings that have taught us something new and that allows us to improve our skills and broaden our horizons. As you can imagine, that’s a lot of books and references, so we have voted for those most relevant to us. In this article we reflect the most recommended books. There are books with a more academic or technical character, but there are also dissemination readings that allow us to better understand some aspects of this exciting world of Data and Artificial Intelligence.

Keep reading!

< Read previous issue

For any question or suggestion, you can also write to

You can enjoy much more content related to data science, innovation, new solutions of financial analytics and how we work in our website:

Let's talk about it. Join the conversation on Linkedin and Twitter.

© 2021 All rights reserved. BBVA AI Factory. Avenida de Manoteras, 44, 28050. Madrid.