Data modelling Proof-of-Concept

This challenge serves as the basis for the technical interview.

Before starting the challenge, please have a look at the challenge and tell us

• if any aspect is unclear to you,
• if you want to amend the challenge to better reflect your skills,
• if you want to change / reduce the scope to better fit your time schedule,
• how much effort in your estimate you will need to finish it, and
• when you can finish the task.

We understand that you have other priorities and can not commit full time to completing the challenge. Thus spend your time economically:

• The task you implement is just a base for the interview discussion. No need to waste your time writing boilerplate code or debugging obscure edge cases.
• If anything is unclear, write us instead of guessing what we expect.
• You don't need to learn a new technology. We can adapt the challenge so that you can solve it with what you already know.

Don't hesitate to write us if you have any questions during the challenge.

Celantur Infrastructure

We have a scalable anonymisation solution with following layers:

• REST API: User uses the REST API to upload images or videos, query the status of the files (new, processing, processed, deleted) and download the anonymised files.
• Storage: Binary files are stored in the object storage (Amazon S3) and processing information (upload-time, processing-time, number of anonymised objects etc) are stored in a relational database (PostgreSQL).
• Processor: Files are processed via stateless processing instances. The instances need to know what objects (face, person, license plate, vehicle) need to anonymised for each image or video.

Tasks

Coding challenge

1. Implement data models for the processor, storage and REST layers using dataclasses or Pydantic.
2. Implement a small REST server using FastAPI with two endpoints (POST /task and GET /task/{id})
   1. POST /task receives the processing information as body parameters, write them into the database and returns the task-id. (Ignore file upload)
   2. GET /task/{id} returns the processing information.
3. Write unit and API tests.
4. Share your (private) repository with @celantor.

Pre-interview preparations

1. Think about how you can implement the Celantur Infrastructure in AWS.
   How do you handle high throughput of files, ensure reliability, security and scalability? What can you concretely do to save costs?
2. Reflect on your last major software project.
   What were the customer requirements?
   What were the (technical and non-technical) challenges?
3. Fill out the Self-evaluation matrix.

Deliverables

• Code.
• Tests.
• Documentation how to run the code.
• Self-evaluation matrix.