Over the last several years I've performed a lot of software engineering interviews. During that time I iterated on the evaluation process to find something that allowed us to make hiring decisions with confidence.
There are a lot of coding interview products and tools out there, but most don't come close to imitating how we work on a day-to-day basis. Whether a candidate can solve a random live coding challenge on the spot isn't an accurate representation of their abilities, especially in a world where we work remotely and asynchronously.
Eventually, I settled on a take-home process of sharing a private Github repo with candidates, asking them to follow the directions in the README and submit a Pull Request - just like we would in our day-to-day work. This gives me a lot of insight into what it's really like to work with that person. For example, I can learn how they communicate in a written and asynchronous manner, how they formulate their commit messages, and how they make decisions.
The downside is that this process was time consuming and manual. For each candidate, I had to take the project source code, create a new project repo, invite the candidate, and keep track of all outstanding projects. So naturally, I created a tool to automate the process :). Feedback and questions welcome!