nasha-firma.od.ua

• https://nasha-firma.od.ua/ - link to the source
• This project is a real website on the Internet, which was carried out for the purpose of training and helping my friends in the development and maintenance of business.
• The project implements the ability to register, authorize and order products from restaurants and cafes. Pagination and multilingual functionality are also implemented. Made some useful features for a specific case
• Tools: Python, Django, PostgreSQL, HTML5, Bootstrap

• This project is implemented using Django Rest Framework. The project implements the possibility of registration, authorization via JWT token and ordering products from restaurants and cafes. Made some useful features for a specific case.
• Tools: Python, DJANGO REST Framework

• I have developed this application as a personal assistant for managing contacts, notes, tags, sending emails from the app to saved contacts, and storing information about them.
• The application automatically fetches news from three news services and renders them on the application's page depends on chosen topic.
• Additionally, a GPT-3.5 Turbo chatbot is integrated into the application, allowing any user to engage in conversations with a virtual bot. The bot tailors its responses based on contact information, including age, gender, name, location, and status.
• Users can also upload, organize, download, and delete files within the service.
• Each user has exclusive access to their own files, notes, and contacts for privacy and security. Logging system using access-token.
Tools: Python, Django, PostgreDB, Dropbox, Aiogram, OpenAI

• Developed a convenient personal chatbot: "LLM_ChatBot,". This innovative tool not only reads and recognizes documents but also features voice recognition for natural interactions. Taking it a step further, "LLM_ChatBot" extracts valuable insights from YouTube videos, leveraging advanced AI methods. Experience a streamlined workflow with this all-in-one solution for document handling, voice commands, and video text extraction.
• It is possible to download data and convert it into text from an Internet link or from a link to a YouTube video, as well as documents in PDF, DOCX, TXT formats. It is also possible to get answers to any questions by talking with GPT-3.5. Text processing and analysis of downloaded documents using the powerful capabilities of Large Language Models (LLM). Saving the text of documents in a vectorized database for later use. Interact with processed documents via chat to get contextually informed answers. Ability to create user profiles, save user data in the database. Security - all users can work only with their documents. In addition, the project provides storage of the history of requests to Large Language Models (LLM).
• The additional level of functionality of the project includes the following possibilities: Using JWT (JSON Web Tokens) for user authentication and authorization. JWT ensures the security of data transfer between client and server by using a signed token. Email verification functionality that allows you to confirm that the specified email address belongs to the user. This may include sending an email with a confirmation link. The ability to reset the user's password by email. The user can send a request to reset the password and receive a link to change the password by e-mail.
• "LLM_ChatBot". This project uses technologies such as: FastAPI, Streamlit, PostgresSQL, Docker, OpenAI, Speech2Text, Faiss, Git, LangChain, Cloudinary, Redis, YoutubeVideoLoader.

• This my cool project from Fastapi

Languages, Libraries and Databases

Python

HTML5

CSS3

MySQL

PostgreSQL

Shell Scripting

Scrapy

Frameworks

Django

Flask

FastApi

Other

Git

AWS

Heroku