thisbra / sepio-classifier

Asset classifier microservice for Sepio code challenge

Sepio Classifier

Asset classifier API microservice project

Architecture

Full view: Figma

Access the Deployed API

https://sepio-classifier.azurewebsites.net/docs

Description

Sepio Classifier is a API microservice built using FastAPI, Pydantic and Clean Architecture with Firebase and Redis for persistence. The knowledge base was imported from the json file using a script to take it to the firestore, where the overrides are also stored in a different collection. Redis is used as a event stream, in this case for reporting inserts on the Override collection. The commits to the main branch will automatically trigger the CD Pipeline and, using Azure container registry, will deploy the container aplication to a Azure Web App. Through Sepio Classifier you can check the Classification and Vendor for a given mac address. To run the application its simple with Uvicorn or Docker, just follow the instructions below. (You'll need a provided .env file)

Project Structure

Running the app

1. Clone the repo or download the ZIP.

2. Place the provided .env file in the root of the project (Where the Dockerfile is).

Option 1 using Docker

You'll need docker installed to run the application.

3. Confirm that you are in the root of the project (/sepio-classifier/.) where the Dockerfile file is and build the image:

docker build -t sepio-classifier .

And wait for the build to finish.

3. Run the app with docker.

docker run --env-file .env -p 8000:8000 sepio-classifier

Option 2 using Poetry and Uvicorn

3. Install the packages described at pyproject.toml

poetry install

4. Activate Poetry's environment

poetry shell

5. Run the API using Uvicorn

uvicorn src.main:app

These commands will start the API that will be available on your machine's ports 8000 (The port needs to be free for the application to run). Now just open your browser of choice and go to http://localhost:8000/docs to test the API. Test with any of the mac addresses in the Knowledge Base json file. And don't forget to test the Override feature!