This is a fully-featured, local RAG (Retrieval-Augmented Generation) chat application. It uses Ollama to run open-source Large Language Models (like Gemma) locally and leverages the Qdrant vector database for efficient semantic search. The entire experience is delivered through a modern web UI. The entire application runs completely on your local machine, ensuring data privacy and security.

StockAgent is an intelligent AI-powered tool that provides real-time stock market insights, news summaries, and data analysis. Leveraging advanced AI and multiple financial APIs, it helps users quickly understand market trends and company news.

Built an AI Agent app that generates real-time stock insights and news summaries using Ollama with open source LLM and OpenAI APIs. Implemented a web scraping tool and Speech to Text feature

The Financial News Analyst is a web application designed to provide users with quick, AI-generated analyses of company earnings reports. By inputting a company name, quarter, and year, users can receive a summarized report of the company's financial performance for that period.

create 3D animation with three.js for mordern website with react framwork and Typescript

Designed and launched a Django-based booking system to simulate online reservation with concurrent booking handling, ensuring conflict-free scheduling. Developed a responsive user interface using HTML, CSS, and Bootstrap, ensuring compatibility across multiple devices. Configured server environment on Linux with Nginx and Gunicorn for application hosting, emphasizing security and scalability. Managed backend operations with a MySQL database, implementing features like user authentication and admin controls.

Developed a model to predict complications from myocardial infarction, improving precision from 0.46 to 0.57 Employed SVM, Random Forest, and TensorFlow multi-layer perceptron models with advanced techniques Achieved model requirements with 0.60 accuracy and 0.77 precision through careful feature selection and threshold setting

Using the Daily Minimum Temperatures in Melbourne dataset which contains data of the daily minimum temperatures recorded in Melbourne from 1981 to 1990. In addition to be using Tensorflow's layers for processing sequence data such as Recurrent layers or LSTMs, also use Convolutional layers to improve the model's performance.

Developed a machine learning model as part of TensorFlow training to predict the next word in a sequence of text involved text preprocessing, creating n-grams, and implementing sequence padding for uniform input size. Designed a neural network using TensorFlow to learn from tokenized text data, and evaluated the model's performance on test data. Demonstrated proficiency in data preprocessing, model design, and TensorFlow framework.

Collaborate with 4 programmers to develop an Android application Grocery List to manage user database as Frontend/Backend Developer finished UI design and login backend activity ahead of schedule by one week.

use C programing language p_thread.h to perform Multi-threaded parallel computing with 12 worker process to read write file with grandchild process(12x3 threads), grandchild calculate array sum, root of sum square array, GEOMETRIC average, get result at same time by pthread_join to get result and save it into file