Smart AI solutions for real-world challenges.
Natural Language Processing
We develop systems that understand, classify, and generate human language.
Computer Vision
Image generation, image classification, object detection, captioning, facial emotion recognition.
Machine Learning
We apply both supervised and unsupervised learning techniques to make predictions or clustering.
Chatbots
AI-powered chatbots that go beyond rule-based flows. By leveraging NLP techniques, we build chatbots that respond in a human-like way.
Smart Address Search
Find real-world addresses from imperfect or informal text inputs. Using NLP and embeddings, we match free-form queries to structured geographic data with high semantic accuracy.
Demand forecasting
Predict future sales or resource needs using historical data. Our models help businesses optimize inventory, staffing, and operations by anticipating demand patterns with precision.
Credit Scoring
Assess the creditworthiness of customers through machine learning models trained on financial and behavioral data. We enable faster, more accurate, and scalable lending decisions.
Customer clustering
Segment customers based on behavior to unlock cross-selling and up-selling opportunities. Our clustering algorithms reveal hidden patterns to support targeted marketing strategies.
Founders
Tomás Ravel
Data Science · Statistics · LLMs
Studied Data Science at the University of Buenos Aires, specialized in Statistics and Large Language Models.
Completed a thesis on turn transition classification in spoken conversations using LLMs and deep learning architectures.
Developed a fully automated trading algorithm at Asesores Inteligentes, achieving a 19% annual return on the QQQ index using models like LightGBM.
Currently teaches AI Engineering at the Universidad de San Andrés, preparing future leaders in artificial intelligence.
Built an intelligent system to compose API workflows from natural language inputs using LLMs and multi-agent architecture, streamlining integrations from unstructured API documentation.
Gonzalo Finkelstein
Data Science · Maths · Computer Vision
Studied both Data Science and Mathematics at the University of Buenos Aires.
At Marvik, led projects involving diffusion-based generative AI including:
- A virtual try-on system for fashion e-commerce.
- A clothing texture designer that applies user-selected fabric textures to clothing designs.
- A leadership detection system for corporate meeting videos, integrating emotion recognition, diarization, transcription, and multi-agent modeling.
Led machine learning initiatives at Meridional Insurance, including a retention prediction model to optimize pricing strategies and increase profitability, as well as NLP pipelines for resume classification.
At Equifax, contributed to the development of credit scoring models, conducted research in Bayesian statistics and machine learning, and implemented robust systems in Python.
Our teams
Machine Learning & Data Science
Leads the development of AI-based systems across a wide range of domains, including traditional machine learning, deep learning, natural language processing, and computer vision.
AI Research Lab
Explores ideas in machine learning and AI to identify and develop the next generation of data-driven products, bridging academic insight with real-world application.
Web development
Is responsible for turning complex AI models into seamless, user-friendly applications.
Gmail: contact@usedataai.com
Linkedin: Use Data