Hey everyone! Welcome back to my Google Summer of Code (GSoC) journey.
In my last post, I shared the story of how I got into open source and was selected for GSoC with NumFOCUS to work on the Neural Network Builder API Refactor project for sbi (Simulation-Based Inference).
Since the official announcement, the past three weeks have been dedicated to the Community Bonding Period. It is designed to help contributors get to know their mentors, understand the community culture, and familiarize themselves with the codebase and tools.
Here is exactly what I did during these past three weeks to get ready for the main coding phase!
The Kickoff Meeting
We started the bonding period with a great kickoff call on Google Meet. It was a joint meeting that included the mentors for both of the selected sbi projects, the selected GSoC candidates. We were also joined by the mentee who successfully completed the GSoC project for sbi last year!
Everyone introduced themselves, and it was incredibly inspiring to meet the team face-to-face (virtually!) and hear about everyone's backgrounds. Having a former GSoC student there was a huge bonus, as they shared some great insights into what to expect in the coming months.
Setting Up the Machine
A big part of getting started is making sure the development environment is properly configured. During our meetings, we discussed the machine setup in detail to ensure both candidates had everything required to run and test the sbi codebase locally without any hiccups.
Embracing AI Coding Assistants
One of the most interesting discussions we had was about using AI coding assistants. In the modern development world, tools like these are becoming standard, and our mentors actually encouraged us to use them!
However, they emphasized using them carefully and strictly following project guidelines. To help us get the most out of these tools without compromising code quality, the mentors shared some excellent Claude code tutorials and provided us with resources to level up our AI coding skills. It is all about using AI to augment our capabilities, not to blindly write code for us.
Diving into the Codebase and Planning
With the setup out of the way, the rest of the three weeks was all about reading. I spent a lot of time diving deep into the sbi codebase and reading through the documentation properly.
Understanding the architecture is critical for my API refactor project. To keep everything organized and on track, we were provided with a plan.md file. This document serves as our roadmap, outlining exactly what needs to be tackled, the skills required, and the milestones we need to hit.
What's Next?
The community bonding period was incredibly helpful. It gave me the time to get comfortable with the team, set up my workflow, and build a solid foundation before the heavy lifting begins.
Now, the official coding phase is here! I am beyond excited to start pushing commits and working on the Neural Network Builder API Refactor.
Thanks for reading, and stay tuned. I will be sharing more technical updates and lessons learned as I start writing code for GSoC!
Top comments (0)