[21 May 2025]
~ I’ve officially been selected for Google Summer of Code 2025!!! In this blog, I’ll break down what GSoC is all about and tell you more about the organization I’ll be working with — Sugar Labs. ~
Firstly, it's not GSocks it's GSoC xD. GSoC or Google Summer of Code is a global program run by Google that offers students the opportunity to work on open source projects during the summer.
If you’re unfamiliar with open source: imagine a project where the code is out in the open for anyone to view, modify, or contribute to. Open source powers a HUGE chunk of the internet today. Without it, developers would constantly have to reinvent the wheel rather than build on existing work—which would be wildly inefficient.
Recognizing the importance of open source, Google created GSoC to encourage more contributions and to foster a more vibrant open source community. Google helps pair contributors with open source organizations and mentors, so they can work together and build cool stuff. Students not only get to work on meaningful real-world projects but also receive a stipend, valuable mentorship, and exposure to global developer communities. It’s a unique opportunity, especially so for students since you are collaborating with experienced developers all aroudnt the globe and learning professional development workflows beyond personal or classroom projects. This was one of the biggest reasons I was excited for GSoC - professional code development and collaboration.
Sugar Labs is a global non-profit organization with a mission close to my heart - improving accessible education for children globally through technology. We're best known for the Sugar learning platform —a playful, activity-based environment that helps young learners explore, create, and collaborate.
Unlike traditional educational tools, Sugar focuses on active learning through activities. It’s a more hands-on experience and encourages kids to think and develop authetic problem solving skills. It empowers them to build projects, solve problems, and learn by doing.
One of the coolest things about Sugar is that it’s open source. Anyone can develop new "activities"(it can be educational mini-app or even a fun game) and add them to their own version of the platform, or better yet, share it with the world! This openness not only allows for innovation but also invites contributions from educators and developers around the world.
At Sugar Labs we believe that every child—regardless of background or geography—deserves access to powerful learning tools. Through our global network of volunteers, teachers, and learners, we continues to imagine, design, and build software that makes a difference in classrooms worldwide.
I’ll be working on modernizing the Speak Activity, it's one of the most beloved activities among children. Speak is an animated face paired with a speech synthesizer. At its core, it speaks out what a student types or says. There’s also a chatbot mode where kids can ask questions and get responses.
Over the summer, I’ll be integrating a Large Language Model (LLM) to enhance both the chatbot and speech features. Some of the improvments / features are:
- Conversational AI: The new chatbot will be powered by a fine-tuned LLM that provides more natural and intelligent responses. It will be able to handle invented spelling and grammatical errors to ensure a smooth and supportive learning experience.
- Natural-Sounding Speech: We’ll introduce a high-quality Text-to-Speech (TTS) model that sounds more natual, with an emphasis on getting pronunciation and phonetics right.
- Offline Support: A lightweight Small Language Model (SLM) will be packaged with the activity as a fallback. This ensures the activity remains useful and accessible for students living in regions with limited or no internet access.
- Cloud Integration: The primary LLM will be hosted on the cloud and accessed via an API. This keeps the local install lightweight while delivering high-quality AI features.
These are some of the very interesting design decisions.
Honestly, I wasn’t even planning to apply. I’d heard of GSoC, but I assumed it was super competitive and that I’d try in my second or third year. But thanks to my senior, SidT, who pushed me to go for it and reminded me I had nothing to lose. So I applied, but it was mostly to better understand how GSoC works, how communities function, how to contribute and how to make a good proposal, so that I could have a better chance at getting in when I did apply during my second or third year. Who knew I would actually make it.
But here is some advice for anyone thats trying to get into GSoC, I felt like doing these helped me:
Contribute to some of their existing repos. This shows your calibre and assures your mentors that you are capable of doing the work. In my case I had even had a small demo of the entire project, so this assured my mentors and myself that I would be able to finish and ship this into the speak activity by the end of the summer.
Communication is key. This is generic advice but it goes a long way, especially since GSoC work is fully remote and communication through online mediums (email, matrix, slack, etc). It's really important you learn how to communicate clearly.
It's okay to not know everything, but it's important you are willing to learn and improve.
Learn to take feedback and make improvements.
I'm absolutely thrilled to have the opportunity to spend the summer building cool things that are used by millions. Unfortunately for me, a decent part of my coding period overlaps with my university final exams (PES why you do this to me -_-), so it's going to be interesting trying to juggle both of them. But I'm very excited and look forward to a summer filled with learnings. I do plan on posting weekly blogs once the official coding period begins, so stay tuned for that!
Powered Not An SSG 😎