Congratulations on passing the initial informational interview! 😊
This next stage is a technical screen, to test your coding ability in a real-world setting.
Task: Create an app/service for people who are blind.
This may be an end-to-end mobile app, or alternatively if you prefer to focus purely on the backend or ML algorithms, you may create a web API which a mobile app would consume instead.
The above requirement is intentionally left open-ended to allow you to use your creativity to solve the problem in whichever way you choose, and to focus on whichever part of the solution matches your skillset best.
As one example, you may choose to create an app which uses the camera to capture images, recognizes text in the image using Microsoft Cognitive Services, and reads it aloud. Another idea would be an HTTP service which accepts an image, does some processing, and returns some JSON. However, do not feel constrained by these suggestion - your app/service may look quite different.
You may choose to use any popular language/framework, but if you are looking for more guidance, the Seeing AI team currently uses C# with Xamarin.iOS for the mobile app, and ASP.Net Core for our backend. However, you may choose to use the technology you are most comfortable with.
Important: Create a file, decisions.md, to document the decisions you make whilst creating the app/service. These may be in terms of what solution to create, technical choices, or any notes for the reviewer.
Please make sure all your code is pushed to GitHub within 5 hours of your allocated start time. The code will be reviewed within one week, and you will be informed whether you will proceed to the final round of interviews with Microsoft engineers.
Good luck, and hope to speak to you soon!