Hiring challenge for the Software Engineer applicants at Radix.
The goal of this challenge is to build and an AI model, found in this repository.
This repository implements a simple scikit-learn model that predicts movie genres based on a given synopsis. In the model folder, you can find the saved model (pickle) as also the python functions created to save, load and make predictions from the model. The data used can be downloaded via the code in the data folder.
To succeed the hiring challenge, you must implement the following:
- Create fastAPI endpoints to make predictions with the model (
/genres/predict). - Create a Docker image to create a container that runs this API locally
Any other best practice implementations are optional, but could be assessed during the technical interview that follows after this hiring challenge. We will also ask you which service you'd use on Azure to deploy it and the steps you'd undertake to deploy. This challenge leaves you the freedom to create a solution that you think would suit our needs, so make sure you can motivate your design decisions!
- Clone this repository
- Ensure that your code is private
- Invite
RadixChallenge(challenge@radix.ai) as a collaborator to your repository
In an environment of choice, install the dependencies using requirements.txt or by running pip install .. Once done, you can run our custom invoke commands, found in the tasks/ folder:
invoke --listto list all the invoke commands (also specified intasks/tasks.py)invoke lintto lint this packageinvoke testto test this packageinvoke serveto serve the APIinvoke runto train and evaluate a model
If all four of the invoke commands run, then you're good to go!
We would love to help you with the challenge, but unfortunately we can't. 😉 That being said, if you find a bug or have troubles setting up your environment, we're happy to help you at challenge@radix.ai!
