[🇧🇷 Português] [🇬🇧 English]

🔐 Cybersecurity, Social Engineering and AI Security / Project 4 – AI Finance Incident Risk & Governance Analysis

Analysis of Algorithmic Bias • Operational Risk • AI Governance Responses in Financial Services

Cybersecurity and Social Engineering Integrated Project - PUC-SP 5th Semester (2026)

Institution: Pontifical Catholic University of São Paulo (PUC‑SP – Humanistic AI & Data Science • 5º Semester • 2026)
School: FACEI – Faculty of Interdisciplinary Studies
Course Repo: INTEGRATED PROJECT: Cybersecurity and Social Engineering – 108 Hours
Professor: ✨ Eduardo Savino Gomes
Extensionist Activities: Extension projects and workshops using open‑source software and data‑driven consulting to support the community, aligned with the 20 official extension hours of the course.

Important

⚠️ Heads Up

• Projects and deliverables may be made publicly available whenever possible.

• The course emphasizes practical, hands-on experience with real datasets to simulate professional consulting scenarios in the fields of Machine Learning and Neural Networks for partner organizations and institutions affiliated with the university.

• All activities comply with the academic and ethical guidelines of PUC-SP.

• Any content not authorized for public disclosure will remain confidential and securely stored in private repositories.

Tip

This repository is part of the flagship project: 🔐 Cybersecurity, Social Engineering & AI Security — Main Hub

Explore the complete ecosystem of materials, analyses, and notebooks in the central repository:

• 🔗 Cybersecurity, Social Engineering & AI Security — Main Hub Repository

*Part of the Humanistic AI Data Modeling Series — where data connects with human insight… and occasionally gets socially engineered. ⚡️

Table of Contents

1. Introduction
2. Objectives and Research Questions
3. Business Context and Data Foundation
4. Methodology — CRISP-DM
5. Data Sources and Preparation
6. Key Variables and Hypotheses
7. Statistical and AI/ML Methods
8. Project Architecture — 5 Notebooks
9. Database Design and REST API Layer
10. Key Findings
11. Timeline, Deliverables, and Business Alignment
12. Installation and Execution Guide
13. Project Structure
14. Limitations and Risk Considerations
15. Conclusion and Next Steps
16. References

1. Introduction

1.1 Business Context

The adoption of Artificial Intelligence (AI) in banking and financial services has expanded significantly across critical domains such as credit scoring, fraud detection, algorithmic trading, risk management, and customer operations.

While these systems improve efficiency and decision-making speed, they also introduce material risks related to model bias, operational failures, and governance gaps. For financial institutions, these risks translate directly into regulatory exposure, reputational damage, and financial loss.

This project analyzes real-world AI incident reports to support a structured understanding of how these risks emerge in financial environments, with a focus on risk patterns, affected customer groups, and governance response effectiveness.

1.2 Business Problem

Given a dataset of AI-related incidents filtered for the financial sector, this project addresses the following business questions:

• Are there recurring risk patterns associated with specific AI use cases (credit, fraud, trading)?
• Do certain customer segments experience disproportionate impact from AI-driven decisions?
• Are governance and regulatory responses aligned with incident severity and risk level?

1.3 Business Value for Financial Institutions

Stakeholder	Business Value
Banks and Financial Institutions	Improved operational risk control and reduced exposure to model failures
Regulators	Data-driven supervision and better risk monitoring capabilities
Risk Management Teams	Enhanced visibility of AI-related operational risks
Compliance Departments	Identification of governance gaps and audit prioritization
Executive Leadership	Better understanding of AI risk impact on business performance and reputation

Note

This project demonstrates how AI incident data can be transformed into actionable risk indicators, predictive models, and API-driven monitoring systems, enabling continuous oversight and improved governance in financial environments.

AI Financial Incident Intelligence System

System Architecture (MLOps Design)

flowchart TB

subgraph DATA_SOURCES
    A1[Kaggle Dataset - Financial Incidents]
    A2[External APIs - Incident Database]
end

subgraph DATA_LAYER
    B1[Raw Data Storage]
    B2[Data Cleaning Pipeline]
    B3[Processed Dataset]
end

subgraph FEATURE_ENGINEERING
    C1[Feature Extraction]
    C2[Data Transformation]
    C3[Feature Store]
end

subgraph ML_PIPELINE
    D1[Model Training]
    D2[Model Evaluation]
    D3[Model Registry]
end

subgraph STORAGE
    E1[(SQLite Database)]
    E2[(Serialized Models)]
end

subgraph APPLICATION
    F1[REST API - FastAPI / Flask]
    F2[Streamlit Dashboard]
end

A1 --> B1
A2 --> B1

B1 --> B2 --> B3
B3 --> C1 --> C2 --> C3
C3 --> D1 --> D2 --> D3

D3 --> E2
B3 --> E1

E2 --> F1
E1 --> F1

F1 --> F2

%% =========================
%% TURQUOISE STYLING (GitHub-safe)
%% =========================

classDef default fill:#0d1117,stroke:#00d1c1,stroke-width:1px,color:#ffffff;
classDef group fill:#0d1117,stroke:#00d1c1,stroke-width:2px,color:#ffffff;

class DATA_SOURCES,DATA_LAYER,FEATURE_ENGINEERING,ML_PIPELINE,STORAGE,APPLICATION group;

Loading

💌 Let the data flow... Ping Me !

🛸๋ My Contacts Hub

────────────── ⊹🔭๋ ──────────────

➣➢➤ Back to Top

Copyright 2026 Quantum Software Development. Code released under the MIT license.