/epsy-backend

Epsy backend technical test

Join Epsy team!

If you want to know more about Epsy have a look at our carriers page, don't hesitate to drop us an email at hiring@epsyhealth.com.

Epsy Backend Technical Test

Background

Our event tracking system generates daily dumps of the data in CSV format stored to S3.

Sample event consist of timestamp, user, event name - example record will look like:

1671192760,"79e0bd79-a40c-4167-b927-44341d27ab81","user.logged"

Additionally, our platform exposes GraphQL endpoint for user data retrieval. API will return user definition object matching technical_test_schema.graphql file.

Both, S3 resource and GraphQL API is protected with username and password (implemented via AWS Cognito).

Task

Create command line tool using click framework which will:

1. Authenticate user using provided credentials against Cognito IDP

TIP

boto3 by default expects AWS Credentials to be provided, that can be changed by setting signature_version configuration parameter to botocore.UNSIGNED

2. Fetch S3 events file from S3 location listed below using boto3 SDK

TIP

You need to exchange Cognito Identity for AWS Credentials - those can be used for communication with S3 This can be done by using:

3. Use user IDs from downloaded file to call GraphQL API and fetch additional information about those users.

4. Provide basic stats about events

  • number of events per day
  • number of events per user role
  • Top 10 users (users with highest number of events) inc. name and last name

Additional requirements:

  1. Fetched additional information should be sent to standard std.out in either JSON or table format
  2. Any errors should be logged to std.error in JSON format

Input data

Cognito Authentication

Client Id: 27s2bt78i19m035mep3aqs4502

IdentityPoolId: us-east-1:6066528c-c7c6-47b9-bd5a-56d2d41afdfe

API

Address: https://v3bdtj5rojdj5okuxsj5lirszi.appsync-api.us-east-1.amazonaws.com/graphql

Schema: technical_test_schema.graphql

Username and password will be provided in separate message.

Please note that our API may occasionally return errors, and your code should handle that.

S3 Data location

Bucket name: hiringapi.epsy.dev

File key: events.csv

Bucket permissions:

  • s3:GetObject
  • s3:ListBucket

Technical requirements

  • Python 3.8+ compatibility
  • Code should be shared via GitHub (public or private repository)
  • Instructions for running the code should be included in README.md file.
  • All exceptions and processing errors should be written to std.err

FAQ

How much time do I have for the task?

There is no time limit, but we strongly recommend to timebox your work to maximum of 2 hours

Do I need to implement all requirements?

Steps 1-3 are required,

May I use stackoverflow, documentation, etc

Yes, all the resources are allowed, as long as you will be able to explain what code is doing

I have more ideas how this code can be improved, but there is not enough time

Feel free to provide all the ideas in your README.md file

Can I share the code with other companies?

Yes, feel free. But please don't share the username details.

Do I need to use listed libraries

Yes, click and boto3 are common libraries in our stack

Useful information

What is important for us during review of your code?

  • Complexity of the code
  • Used libraries
  • Consistency across the codebase
  • Compatibility with Python standards
  • Cleanness of the code
  • Error handling and error reporting/logging
  • Performance
  • Memory consumption
  • Output formatting
  • User experience - that tool can be potentially used by non-devs
  • Security
  • Documentation