Experience

Uali

• Argentina
• Technology, Information and Internet
• 1 - 100 Employee

• Data Scientist

  • Jun 2023 - Present

  Uali is a technology-driven company that integrates robotics, IoT, and artificial intelligence to provide simplified monitoring services, optimizing time and cost for clients while minimizing environmental impact. I have had the privilege to work on several key projects: Analog to Digital Gauge Measurements Conversion: Developed algorithms for the precise conversion of analog gauge measurements into digital format. This breakthrough enabled automatic gauge reading, ensuring seamless functionality and operational integrity of the gauges. Vehicles Near Assets Detection: In this initiative, we trained and evaluated machine learning models for detecting vehicles in close proximity to valuable assets. This bolstered security measures and enhanced safety protocols for clients. Gate Status Detection: By implementing an automatic gate status detection system using computer vision techniques, this project contributed to accurate monitoring and provided valuable insights to customers. Show less



Anyone AI

• United States
• E-Learning Providers
• 1 - 100 Employee

• Machine Learning Engineer

  • Feb 2023 - Present

  AnyoneAI is a highly selective fellowship program, and I feel privileged to be part of it with less than 2% acceptance rate. This opportunity has allowed me to collaborate on diverse projects and expand my expertise in Machine Learning Engineering: *Financial Advisor Chatbot: Leading a team of four, we built a powerful Question-Answering chatbot utilizing LLMs like GPT 3.5. The chatbot can provide up-to-date information on the financial status of over 100 NASDAQ companies, processing a dataset of approximately 1000 PDF documents. *Movie Review Sentiment Analysis: Implementing sentiment classification on more than 50,000 movie reviews, we utilized techniques such as TF-IDF, Word2Vec, and Logistic Regression, achieving an impressive 0.96 ROC AUC score on testing. *Vehicle Image Classifier: Conducting image classification on a dataset of vehicle images for 25 different classes, we achieved an accuracy of over 89% on testing by training a CNN using transfer learning with ResNet and EfficientNetB0. Additionally, background removal techniques were implemented using Detectron2, further enhancing the model's performance. *Home Credit Risk Analysis: Conducted risk analysis for home credit applicants, processing and manipulating a dataset with over 240,000 rows. We achieved a remarkable ROC AUC score of over 0.77 using supervised models like Logistic Regression, Random Forest, and LightGBM. Show less



CONICET

• Argentina
• Government Administration
• 700 & Above Employee

• PHD Student

  • Apr 2021 - Feb 2023

  Implementation of neural network models and machine learning techniques, such as dimensionality reduction techniques and unsupervised clustering, for the optimal design of T-cell chimeric antigen receptors for cancer treatment. The dataset generated and used was composed of several curated coarse grained molecular dynamics. Implementation of neural network models and machine learning techniques, such as dimensionality reduction techniques and unsupervised clustering, for the optimal design of T-cell chimeric antigen receptors for cancer treatment. The dataset generated and used was composed of several curated coarse grained molecular dynamics.



Comisión Nacional de Energía Atómica

• Argentina
• Government Administration
• 400 - 500 Employee

• Intern

  • Jul 2019 - Feb 2021

  Identification of biomarkers based on fluorescent proteins with non-invasive magnetic resonance imaging. I performed data analysis with Python of medical images obtained in magnetic resonance imaging experiments that I designed and implemented. Identification of biomarkers based on fluorescent proteins with non-invasive magnetic resonance imaging. I performed data analysis with Python of medical images obtained in magnetic resonance imaging experiments that I designed and implemented.