GitHub Copilot – And the impact on my big project - Part 2

In the first part I talked about how GitHub Copilot had helped me with a lot of supporting tasks. What follows now is the story of how big of an impact it had on a project that needed a lot of code writing and where I was the only person writing code on that project.

When looking back at the beginning of the project I did not have GitHub copilot at my side, but a few weeks into the project I gained access to it, and it made me more efficient instantly! The project was about setting up an integration between our HR system and all other systems that deepened on HR data.

We were moving away from an on-premises solution that was based on an SQL database and the all the existing integrations were built to “talk to” an SQL database. The new HR system is a SaaS solution and had no SQL database and only an API. So, to save time, I decided to build a new SQL database and use the API the new HR system provided to load data into that new database. As mentioned, I didn’t have GitHub Copilot in the beginning but looking back I could “feel” that it was missing because the first code that interacted with the API to load data into database took quite a while to setup.

This is where GitHub copilot came into the picture. I had to write (literary) thousand lines of code to get the data from the API into the database and sort it into different tables. It was a big task to do on my own (but I knew I could do it on my own, it was just a matter of time), but the GitHub copilot saved me countless hours of writing code and debugging. A lot of time that would been spent on the internet searching for help on a specific error or how to accomplices different task was now gone, and the help to figure out how to deal with error handling in a specific part of the code or how to optimize specific parts to make the code run faster. After a few weeks of working with GitHub copilot I was done setting up the database and finishing of the integration to other systems.

Now the turn had come to our IAM system, which we replaced at the same time as we replaced our on-premises HR system for SaaS HR system. We said goodbye to our IAM system that was managing everything related to user management (onboarding, offboarding, movers etc.) It is now going to be replaced by Microsoft Entra ID Governance. The first order of business was to be able to create users, using the data from the new SQL database and make Microsoft Entra ID manage the process, which needed some coding as well. A task that GitHub copilot also helped save a lot of hours, the same goes for the offboarding process and the whole handling of employees being promoted or change department etc.

While Microsoft Entra ID Governance provides a lot out of the box, there is also a lot of custom coding that needed to be done for it to fit into “our way of life”. Once again, the GitHub copilot was very useful when I wrote the code that was required to adapt it to the process in my company. In the end, I honestly think that I did the work of 2 people with the help of GitHub. Because writing code was one thing, but in helping me create documentation for all of it was helpful as well.

I want to sign off by saying that GitHub copilot has changed my way of writhing code and made me more productive, and it was scary how fast I got “addicted” to it. It really is a new way of working.