CHAT + OPEN AI API
CHAT + OPEN AI API
Embedded Files
2023 / 2024
2023 / 2024
ABOUT
ABOUT
The project began with the client's need to develop a chat to help users identify hate speech based on transcripts extracted from media.
Therefore, the chat was developed to be incorporated into an existing system made in React JS and back-end in Node.js. Where communication between the system and chat was done through post messages.
The chat was developed in Flutter Web, due to customer demand, and the back-end in Python. Communication between front-end and back-end was done via websocket. File storage done by vectorstore using QDrant. Chat history managed through caching by Redis and data storage with a MySQL database.
The chat communicated with the OpenAI API, which returned streams with the requested information to the user.
DAY TO DAY PROJECT
DAY TO DAY PROJECT
Daily meetings on weekday.
Sprints of 2 weeks long.
Refinements twice a week.
Retrospectives every sprint end.
Planning every sprint start.
We used kanban to manage user histories and epics.
TECHNOLOGIES
TECHNOLOGIES
Google Sites
Report abuse