PROJECTS
Aneurysm Diagnostic Assistant
Developed a RAG-based tool with OpenAI GPT-4 and FAISS to assist medical professionals and researchers in extracting insights from medical literature, enhancing efficiency and decision-making in aneurysm management.
RAG Application for Disarm Framework
Designed a RAG application for the Disarm Framework using GPT-4o and Llama2 Uncensored, implementing FAISS and vector embeddings to provide secure, contextually accurate data retrieval.
Vendor Food Supply Management using RAG
Developed a RAG application for Vendor Food Supply Management to process and analyze vendor data from diverse PDF formats. Key features include data extraction and translation from Italian to English using MarianMT, structured JSON data storage in MongoDB, and real-time querying with OpenAI's GPT-4 for advanced insights.
Project Tampere city
Innovative AI-driven application for all your questions about the Land Use and Construction Act of Tampere, Finland. This RAG App provides accurate answers in Finnish and English using OpenAI API and GPT-4 LLM model, easy to navigate urban development rules.
Chat2Query
An LLM project leveraging Google Palm and Langchain to develop a question-answer system for a T-shirt retail store. The system translates natural language queries into SQL, executes them on a MySQL database, and generates responses. Key components include Google Palm LLM, Hugging Face embeddings, Streamlit UI, Langchain framework, ChromaDB vector storage, and few-shot learning techniques.
Restaurant Search Menu Generator
This project is a unique blend of OpenAI's Large Language Model (LLM), Streamlit, and the LangChain framework. The Restaurant Search & Menu Generator is a simple yet powerful tool that uses artificial intelligence to generate restaurant names and menu ideas.
URL: https://github.com/toufique69/Restaurant_Search_Menu_Generator
Hospitality Dashboards
Microsoft Power BI project in the hospitality domain. Used DAX, and Power Query to develop a comprehensive Business and Data Intelligence Dashboard. Created custom metrics and delivered actionable insights to support revenue optimization.
E-Commerce Sales Dashboard
Conducted a comprehensive analysis of e-commerce sales data and subsequently designed an interactive dashboard employing Microsoft PowerBI.
URL: https://github.com/toufique69/E-Commerce_Sales_Dashboard
CNN-RNN-MLP
- MNIST handwritten digits classification with MLPs and CNNs
Train a multi-layer perceptron model to classify MNIST digits and train a convolutional neural network (CNN, ConvNet) to classify MNIST digits using TensorFlow with the Keras API.
- IMDB movie review sentiment classification with RNNs
Train a recurrent neural network (RNN) for sentiment classification using Tensorflow with the Keras API.
Frequent Itemsets and Text Similarity Search with MinHash
Frequent itemset analysis using FP-Growth and text similarity search using MinHash. It includes scripts for analyzing the MovieLens dataset and performing similarity search on text documents.
URL: https://github.com/toufique69/Frequent_Itemsets_and_Text_Similarity_Search_with_MinHash
Page Rank Algorithm and Social Network Graph
Data mining exercise involving graph analysis and social network clustering. The exercise uses the SNAP social circles dataset and covers topics such as graph creation, PageRank algorithm, community detection, and visualization.Data mining exercise involving graph analysis and social network clustering. The exercise uses the SNAP social circles dataset and covers topics such as graph creation, PageRank algorithm, community detection, and visualization.
URL: https://github.com/toufique69/Page_Rank_Algorithm_and_Social_Network_Graph
Movie Recommendation System
User-based and Item-based Collaborative Filtering Recommendations approach has been implemented to provide movie recommendation to the users using the Pearson Correlation Function.
URL: https://github.com/toufique69/Movie-Recommendation-System
Group Recommendations
Producing group recommendations using the user based collaborative filtering approach.
URL: https://github.com/toufique69/Group-and-Sequential-Recommendations
Patient Condition Prediction
Finding laboratory results of patients and the label whether he/she has a fatal condition or not using K-Means clustering for two dimensional and also for three-dimensional data.
URL: https://github.com/toufique69/Patient-Condition-Prediction
Shopping Store
t's a desktop application and the goal of this software is to keep the data safe and also a user-friendly experience for selling.
Platform: Windows, Language: Java, Database: Oracle 10g/Microsoft SQL Server 2014/Microsoft Access, Software: NetBeans IDE 8.0.2.
Chat Messenger
A sample chat messenger using Java programming language.
Healthcare BD
Platform: Windows, Language: C#, Database: MS SQL Server 2014, Software: VS-2015.
InfoBD.com
It's a website that provides all kinds of information, location, address etc. about POLICE, RAB, Fire Services of Bangladesh.
Platform: Web-based, Language: ASP.NET, Database: MS SQL Server 2014, Software: VS-2015.
Becha-kena.com
Becha-kena.com is an online e-commerce site. It provides consumers with quality and demandable products.
Platform: Web-based, Language: PHP, HTML, CSS, JavaScript, Database: MySQL, Software: XAMPP, Sublime Text.
Racing Game - Run
It's an android racing game playing with sensors.
Platform: Android, Language: Java, Software: Eclipse IDE.
Snake and Ladder Game
Platform: Windows, Language: C++, Software: VS-2010.
History of Bangladesh, Hospital Bangladesh, Sports News
Three different types of Windows Phone App.
Platform: Windows Phone, Language: C#.
Microcontroller Based Line Following Robot
Items Used: Arduino UNO, IR Sensor, Motor Shield, Lithium Battery.