franspg Goto Github PK

VU Security

• 📉 Detection and interpretation of fraudulent transactions with a machine learning approach using FastApi, scikit-learn, XGBoost, LIME, Celery, Pandas, Numpy.
• 🔬 Gathering images from the internet massively using Selenium, FastApi, Celery.
• 👨 Face recognition, face verification, face alignment, etc. Using Azure Cognitive Services, DeepFace, among others.

ElevenPaths, Telefónica Tech

• 🔎 PackageDNA, an open source tool that gives developers, researchers and companies the ability to analyze software packages of different programming languages that are being or will be used in their codes, providing information that allows them to know in advance if this library complies with processes.
• 📰 DIARIO scans and analyzes documents in a static way keeping users safe against virus and threats, guaranteeing users and their documents content private.
• 🎆 Aldara, a social network intelligence framework that helps analysts extract information from social media interactions.

LatinMalls

• 💵 Engine for AdWords accounts automatization and bids budget optimization.
• 🔧 real-time gathering tool for Google Ads API.
• 💬 NLP to block and generates new keywords for the main engine.

Universidad de Palermo

• ♻️ Weather sensor fault detection in meteorological masts

📝 Recent blog posts

🇪🇸 Spanish

• Cómo engañar a las apps que usan Deep Learning para la detección de melanomas
• Las primeras vulnerabilidades oficiales en Machine Learning, así, en general
• Pensando en ataques a WAFs basados en Machine Learning
• Adversarial Attacks, el enemigo de la inteligencia artificial (II)
• Adversarial Attacks, el enemigo de la inteligencia artificial

🇪🇸 Spanish Personal blog

• Generación de datos artificiales (Data Augmentation)
• Inteligencia artificial para contrarrestar a la inteligencia artificial (DeepFake)
• Ciencia de datos aplicado a la astrofísica
• Analizando los subtes de Buenos Aires