Giuliano Ojeda

Materials for the LLM Engineering Essentials course

Policy Pro is an LLM agent 🤖 specializing in insurance policy analysis. It features a chatbot interface built with Streamlit and is powered by OpenAI models (like GPT-4). The agent is fully containerized with Docker for easy local setup and cloud deployment.

A serverless API 🛰️ that provides unique device counts for a given geographic location. Built with Python on AWS Lambda, it uses Uber's H3 geospatial indexing for efficient querying. The application is deployed using the AWS SAM CLI and securely manages credentials with AWS Secrets Manager.

A containerized image classification service 📦. Uses a pre-trained TensorFlow CNN served by a Python/Flask API, with a simple web UI for uploads.

The official code repository for the MIT Introduction to Deep Learning course 🧠. This repository contains all software labs, which are Jupyter Notebooks designed to be run on Google Colab with GPU acceleration. These hands-on labs complement the public lectures and utilize the custom mitdeeplearning Python package to teach the fundamentals of deep

An image classification project 🚗 to identify a vehicle's make and model from a picture. This is a multi-class classification task solved using Convolutional Neural Networks (CNNs) built with TensorFlow and Keras. The workflow is guided by a Jupyter Notebook, with support for Docker and Google Colab for GPU-accelerated training.