About
I'm Othman SAMIH
As a second-year master's student in Data Science (Specialization in Artificial Intelligence & Management) at IA School, I'm driven by a passion for AI and machine learning. My focus is on mastering cutting-edge techniques and applying them to real-world projects.
Projects
WebGPT
An interactive web search assistant combining OpenAI's GPT models and the DuckDuckGo Search API to deliver concise, accurate responses. It features advanced tools for searching text, images, videos, and news, presented through a user-friendly chat interface built with Streamlit.
RAG-Enhanced-Chatbot
A sophisticated Retrieval-Augmented Generation chatbot for interactive, document-based conversations. It uses embeddings, vector databases, and an AI-powered summarizer to deliver accurate and context-aware responses.
Agents Powered Newsletter
A streamlined solution for automating the research, editing, and formatting of personalized newsletters. It features a modular architecture for content gathering, editing, and HTML newsletter assembly, complemented by a user-friendly Streamlit-based GUI for seamless interaction.
AI Marketing Team Automation
A project that automates researching, planning, and creating Instagram content using agent-based architecture and predefined tasks, streamlining the content creation process for Instagram pages.
Meeting Preparation Automation
A streamlined solution utilizing intelligent agents to research, analyze industry trends, develop strategies, and generate concise briefing documents, ensuring well-informed and productive meetings.
PDF Insights Processor
AI-Powered PDF Insights: A Python project that extracts, processes, summarizes, and visualizes data from PDFs, enabling question-answering with context-aware AI models.
Coffee Shop Customer Service Chatbot
An AI-powered chatbot for a coffee shop app, leveraging LLMs and NLP to manage orders, answer menu queries, and provide personalized recommendations. It uses a modular agent-based architecture with specialized agents and a guard to filter irrelevant queries.
Road risk detection
A real-time system using vehicle-mounted cameras to detect speed bumps and potholes, estimating their distance to enhance driver awareness and safety. It leverages computer vision techniques to alert drivers to potential road hazards.
Moroccan law chatbot
Moroccan law chatbot providing answers on various legal topics including business, taxes, and divorce, using a cloud-based dataset of Moroccan law documents.
Real-time UK license plate detection and recognition
A real-time UK license plate identification system using advanced computer vision and machine learning techniques for accurate detection and character recognition.
Segmented bottles in a production line
Designing and implementing a cutting-edge system for segmenting bottles on a production line, leveraging the power of YOLOv8.
Deployed ML model as Dockerized API
This project streamlines the deployment of machine learning models by creating a Dockerized API. This makes the model accessible and usable as a web service, allowing others to easily interact with it and get predictions.
QR code attendance system
A QR attendance system designed to manage access control by scanning QR codes, providing clear indications of whether access is accepted or denied.
Object Detection with AWS Rekognition
A system that uses Amazon Rekognition Object Detector to identify objects in images and videos. It can detect a wide range of objects.
Pneumonia classification web app
A pneumonia classification web app that identify pneumonia in chest X-ray images.
Multiple disease prediction system
A predictive system capable of forecasting multiple diseases, including Diabetes, heart disease, and Parkinson's, through the application of Machine Learning techniques.