wavelab_challenge

NOI HACKATHON wavelab_challenge

Intro

This end-to-end Docker Compose application aims to provide a useful recreational opportunity for users waiting for their electric vehicle to be fully charged. It uses up-to-date data from the [Open Data Hub (ODH)] (https://opendatahub.it) to optimise the OpenAi GPT-4o through Retrieval-Augmented Generation (RAG) in order to provide suggestions based on the user's preference, availability of events, activities, weather and distance between the user and the point of interest. The application uses the Flask framework to provide optimal interaction.

Installation

This Docker Compose application involves building custom images for its Python processor components.

Clone this repository and copy .env.example into .env, possibly editing it (just creating .env is enough):

$  git clone https://github.com/giusber2005/wavelab_challenge.git
$  cd wavelab_challenge
$  cp .env.example .env

Usage

In a terminal, use Docker Compose to start or stop all the required components.

$  docker compose up                        # to start the application
$  docker compose down -v --remove-orphans  # to stop the application

Once the application is up and running, you can access the chat and start your query by:

• Open your web browser and go to http://localhost:5000.
• You will be presented with a flask web interface showing the chat, the platform is already designed to work on mobile too.
• The data is updated at the first prompt retrieving the daily atraction and the weather forecast for the next three hours.
• Explore the different options by continuously chatting with the boot and specifying all your needs.

Architecture

flowchart LR
    odh("OpenDataHub API"):::compext
    openai("OpenAI API"):::compext
    ui("browser showing\nthe HTML"):::compext
    subgraph app [Docker Compose application]
        flask("Flask"):::comp
        sqlite3("SQlite3"):::comp
        frontend("Flask frontend server"):::comp
    end
    odh --> flask
    openai --> flask
    flask --> frontend
    frontend --> ui
    flask -.- sqlite3
    classDef scope fill:#fff,stroke:#333,stroke-width:1px,stroke-dasharray: 5 5,color:#444,font-size:10pt;
    classDef comp fill:#fafafa,stroke:#333,stroke-width:1.5px,font-size:10pt;
    classDef compsub fill:#eee,stroke:#333,stroke-width:1.5px,font-size:10pt;
    classDef compext fill:#fff,stroke:#333,stroke-width:1.5px,font-size:10pt;
    classDef none fill:#fff,stroke:#fff,stroke-width:0px,font-size:0pt;
    class app scope

Loading

The figure shows the application architecture in terms of components and data flow (solid links):

The application uses current data from the [Open Data Hub] (https://opendatahub.it). Loads the :

• Weather forecast for the next 3 hours
• Activity based on the season
• Events happening during the day
• Distance between EV charging station and POI

The information is collected and stored to create a data set used to run RAG with the GPT-4o model. The props for the LLm are tailored so that the models are limited in the hallucination, and do not drive away from the conversation. The conversation is stored in a SQlite database to further fine-tune the model with enough user preferences.

Licence

MIT