vtc - Uber Ride Fee Prediction

This project is aimed at predicting the fee of a Uber ride using machine learning techniques. As a classical project, it contains:

• training/exploration repo for ML and datascience
• app repo with functions to deploy the ML model and associated unit tests. The code of this repo is formatted by a linter.
• devops repo for potential CI/CD
• infra repo to show an example of infrastructure as a code (IAC)

Get started!

1. Training

Repository

• The training directory contains a Jupyter Notebook (model_training.ipynb) where the machine learning model is developed and trained.
• To reproduce the model or experiment with different data, navigate to the training directory and follow the instructions in the notebook.

2. App

The App is designed as following:

Stage 1: Trigger

• Description: This stage serves as the initial trigger point for the workflow.
• Trigger Event: It can be initiated by an event such as a user's action (e.g., requesting a ride) or through a scheduled job.

Stage 2: Put in Queue

• Description: After the trigger event, the workflow places the task or job into a queue.
• Purpose: Queues are used for managing asynchronous tasks and ensuring efficient processing.

Stage 3: Trigger Second App

• Description: This stage trigger the second application that makes the real process.

Stage 4: Call External API to Get Meteo Data (Not developed yet)

• Description: In this stage, the workflow makes an HTTP request to an external API to retrieve meteorological (weather) data.

Stage 5: Read Model into Data Lake

• Description: This stage reads the model into the data lake.

Stage 6: Send Callback

• Description: The workflow concludes by sending a callback or notification.

Azure App

• The app directory includes an Azure web application designed to make predictions based on the trained machine learning model.
• You can deploy this app on Azure to provide a user-friendly interface for predicting Uber ride fees.

3. DevOps

Continuous Integration and Continuous Deployment (CI/CD)

• The devops directory contains the necessary configuration and scripts for implementing CI/CD for the Azure app.
• We use Azure DevOps to automate the build and deployment processes.

4. Infrastructure

Infrastructure as Code (IaC)

• The infra directory holds the infrastructure as code (IaC) scripts that define the resources needed to run the Azure app.
• These scripts can be used with tools like Terraform to provision and manage the required infrastructure.

Getting Started

Follow these steps to get started with the project:

1. Clone this repository to your local machine, preferrably on VSCode:

   git clone https://github.com/C-BdB/vtc.git
   

2. Install dependancies of the requirements.txt

3. Install Azurite Extension on VSCode, Azure Functions Core tools

4. Use Postman to send requests (see the body of a typical request in the app/api_definition repository

5. Prepare a webhook to receive the asynchronous response

6. Run the code

   func host start