Discover Feedzai Research
Applied Research to produce the best financial risk mitigation product in the world.
We don’t just build models. We nurture the environment in which they can adapt faster, defend customers more robustly, and instantly reveal shadows in the global financial network.
Recent Publications
The Balance between Nuance and Clarity: Decluttering Tabular Sequential Graphs to Counter Money Laundering
Money laundering is not only about moving illicit funds, but about hiding the money’s origin and traces to complicate detection. Financial criminals resort to many methods to avoid regulators and legal thresholds. But analysts investigating alerts, dedicated to pin mule accounts and track suspicious transactions daily, also have theirs. Network visualizations can be key in countering adversarial money laundering activities, especially if they provide a clear overview of the money flows and a seamless analysis experience, but they are often not structured for this type of task. That is why we propose a tabular sequential graph visualization tailored to money laundering analysis – following transactions (edges) from the victim account that triggered an alert through multiple accounts (nodes) and banks (rows). To reduce the number of nodes and edges, we propose three methods for grouping these tabular sequential graphs: an amount-based approach, a time-based approach, and a combined solution that considers both the transaction amount and its order. A user study with experts revealed that the most effective method in node reduction was not necessarily the most interesting for analysis and that there is a trade-off between manual work and time for interpretation in more granular graphs.
Uncertainty-Aware Systems for Human-AI Collaboration
Learning to defer (L2D) algorithms improve human-AI collaboration (HAIC) by deferring decisions to human experts when they are more likely to be correct than the AI model. This framework hinges on machine learning (ML) models’ ability to assess their own certainty and that of human experts. L2D struggles in dynamic environments, where distribution shifts impair deferral. We argue that robust HAIC in dynamic environments requires uncertainty-driven policy switching rather than reliance on a single deferral strategy. To operationalize this principle, we introduce two uncertainty-aware approaches that estimate epistemic uncertainty to guide the deferral policy choice. Both methods are the first uncertainty-aware approaches for HAIC that also address limitations of L2D systems including cost-sensitive scenarios, limited human predictions, and capacity constraints. Empirical evaluation in fraud detection shows both approaches outperform state-of-the-art baselines while improving calibration and supporting real-world adoption.
Responsible AI & The TRUST Framework
Why bother with Responsible AI? What are the tradeoffs? How to get started with Responsible AI? What is the TRUST Framework? In this keynote Pedro presents some context and some misconceptions about Responsible AI and shows how the TRUST Framework can guide the development of Responsible AI systems.
SARSum: A Relevance and Comprehensiveness-Aware Abstractive Summarization Dataset for Suspicious Activity Reports
Existing benchmarks that evaluate the ability of Large Language Models (LLMs) to summarize rely primarily on measuring a summary’s lexical similarity to a reference or on assessing whether its claims are factually consistent with the source document.
Benchmark It Yourself (BIY): Preparing a Dataset and Benchmarking AI Models for Scatterplot-Related Tasks
AI models are increasingly used for data analysis and visualization, yet benchmarks rarely address scatterplot-specific tasks, limiting insight into performance. To address this gap for one of the most common chart types, we introduce a synthetic, annotated dataset of over 18,000 scatterplots from six data generators and 17 chart designs, and a benchmark based on it. We evaluate proprietary models from OpenAI and Google using N-shot prompting on five distinct tasks derived from annotations of cluster bounding boxes, their center coordinates, and outlier coordinates. OpenAI models and Gemini 2.5 Flash, especially when prompted with examples, are viable options for counting clusters and, in Flash’s case, outliers (90%+ Accuracy). However, the results for localization-related tasks are unsatisfactory: Precision and Recall are near or below 50%, except for Flash in outlier identification (65.01%). Furthermore, the impact of chart design on performance appears to be a secondary factor, but it is advisable to avoid scatterplots with wide aspect ratios (16:9 and 21:9) or those colored randomly. Supplementary materials are available at https://github.com/feedzai/biy-paper.
3 Research Areas. 1 Mission.
We don’t just build models. We nurture the environment in which they can adapt faster, defend customers more robustly, and instantly reveal shadows in the global financial network.
Feedzai Reaches 100 Filed Patents
Awarded People
Catarina Belém
IST Maria de Lourdes Pintasilgo award, 2021; Fulbright Scholarship, 2021
André Cruz
Outstanding MSc Thesis Award for the best thesis in "Computational Intelligence" by IEEE Portugal, 2020
Sofia Gomes
Portuguese Women In Tech Awards finalist in the Systems and Network Engineer category, 2021
Research in Action
Explainers from the researchers solving tomorrow's problems today.
Finding the Truth: No More Compromises
Feedzai’s new engine loves data, the more the better. It enables our AI to find answers in the data that […]
Responsible AI: Identify and Address AI Biases (Feedzai Spotlight Sessions)
Thanks for watching our latest Spotlight Sessions on responsible AI! For more around artificial intelligence in banking and risk management, […]
Latency in Fraud Prevention: Don’t Let the Numbers Fool You (Transaction Latency Explained)
If a fraud solution provider is boasting about their latency numbers, be cautious—it’s easy to be misled. Latency is simply […]
The Feedzai Research Blog
Home to an ongoing series of exciting tales and cool articles, elucidating the latest developments on how our experts combat financial crime through continuous innovation in data visualization, system research, engineering, data science and Artificial Intelligence.
Benchmark It Yourself (BIY): Preparing a Dataset and Benchmarking AI Models for Scatterplot-Related Tasks
When we need to visualize and interact with millions, or even just thousands, of individual points while analyzing data, we typically resort to rendering them in the browser using a canvas. The other common approach for the web, SVG, doesn’t scale when the number of individual elements increases to such quantities. However, while solving one problem, canvas charts introduce a new challenge: accessibility.
Benchmarking LLMs in Real-World Applications: Pitfalls and Surprises
Moving beyond binary classification provides novel insights.
Causal Concept-Based Explanations
Over the years, we have evolved from using simple, often rule-based algorithms to sophisticated machine learning models. These models are incredibly good at finding patterns in large datasets, but due to their complexity it is frequently challenging for a human to understand why a certain input leads to its respective output. This is especially problematic in areas where high-stakes decisions are being made and where human-AI collaboration is critical.
Feedzai TrustScore: Enabling Network Intelligence to Fight Financial Crime
Detecting financial fraud is like finding a moving needle in a shifting haystack. Fraud accounts for a tiny fraction of financial transactions, often less than 0.1%. At the same time, fraudsters are constantly adapting their tactics to evade detection. And this happens within a live and dynamic environment, where financial behaviors and technologies are changing over time. In short, this is an exceptionally difficult problem for financial institutions.
Paying it Forward:
Open-Source at Feedzai
We give back to the community that makes our work possible. Our open-source packages are used by data scientists and ML engineers worldwide and have had over 1M+ downloads.
Aequitas
A bias auditing and Fair ML toolkit for data scientists, machine learning researchers, and policymakers. Aequitas was started by the Data Science for Social Good Foundation.
FairGBM
An easy-to-use and lightweight fairness-aware ML algorithm with state-of-the-art performance on tabular datasets.
AutoVizuA11y
A React library to automate the creation of accessible data visualizations.
FiFAR
A Fraud Detection Dataset for Learning to Defer.
Academic Partnerships
Feedzai Research academic partnerships span internships, scholarships, PhD supervision, and university partnerships.
- ~30 MSc and PhD theses supported or co-supervised
- Sponsors of the Feedzai Chair of Machine Learning, currently held by Professor Mario A. T. Figueiredo
- Sponsors of the Women in Science scholarship program at the University of Lisbon, supporting over 15 MSc and PhD students since 2022
- Partnership with the University of Coimbra to sponsor the First Foundation Project
Recent Awards and Analyst Recognition
World Changing Ideas - Software category finalist, 2021
Fast Company - World Changing Ideas - AI & Data category finalist, 2021
Fintech Breakthrough Awards - Fraud prevention Innovation of the Year, 2021
Asia Fintech Awards - Regtech of the Year, 2021
The Stack - Tech for Good, 2021
Page printed in June 15, 2026. Plase see https://www.feedzai.com/research for the latest version.