My Google Summer of Code 2025 Journey with the Scala Center

By Gopi Trinadh Maddikunta

The summer of 2025 was a turning point in my open-source journey. I had the incredible opportunity to be selected as a Google Summer of Code (GSoC) 2025 contributor with the Scala Center at EPFL — and today, I’m proud to share that I’ve successfully completed the program.

How It All Started

As a Master’s student in Engineering Data Science at the University of Houston, I was already deep into the world of retrieval-augmented generation (RAG), embeddings, and vector search. But I wanted to take my skills beyond coursework and research — I wanted to contribute to something real, something that thousands of developers would use.

That’s when I came across the Scala Center’s GSoC project list. One project immediately stood out: building a robust Scala interface for Retrieval-Augmented Generation (RAG). It was the perfect intersection of my expertise in information retrieval and my interest in strongly-typed, functional programming ecosystems.

I reached out to my mentors, wrote my proposal, and after a nerve-wracking wait — I was in.

The Project: Bringing RAG to the Scala Ecosystem

The goal was ambitious: bring the power of RAG into the Scala ecosystem by building a production-grade Scala interface that developers could use to integrate retrieval-augmented generation into their applications.

What I Built

Over 12+ weeks of focused development, I worked on:

• A Scala-native RAG interface — Designed and implemented a clean, idiomatic API that lets Scala developers plug into RAG pipelines without leaving the ecosystem.
• Embedding and vector search integration — Built connectors for embedding models and vector stores, making similarity search first-class in Scala.
• Evaluation-driven development — Every component was benchmarked and tested rigorously. I didn’t just build features — I proved they worked under real-world conditions.
• Documentation and examples — Because code without documentation is just a private experiment. I made sure every module came with clear guides and runnable examples.

Technical Highlights

Working in the Scala ecosystem meant embracing functional programming patterns, type safety, and composability. Some of the most rewarding technical challenges included:

• Designing APIs that felt natural to Scala developers while wrapping complex ML operations under the hood.
• Balancing latency and throughput — RAG systems need to be fast, and I spent considerable time profiling and optimizing the retrieval pipeline.
• Making the system evaluation-driven from day one, so every change could be measured against real benchmarks rather than gut feelings.

What I Learned

GSoC taught me far more than just code:

• Open-source collaboration is a skill. Writing code that others will read, review, and extend requires a different level of discipline. Clear commit messages, thoughtful PR descriptions, and responsive communication became second nature.
• Mentorship matters. My mentors at the Scala Center provided guidance that went beyond code reviews. They helped me think about API design, long-term maintainability, and the broader ecosystem impact of my work.
• Shipping is different from building. Getting something working on my machine was step one. Getting it to a state where the community could adopt it — with documentation, tests, and clean APIs — was the real challenge.

The Bigger Picture

This project wasn’t just a summer internship. It was about making advanced AI techniques accessible to an entire programming community. The Scala ecosystem has always been strong in data engineering and distributed computing, but RAG support was a gap. I’m proud to have helped fill it.

For me personally, this experience reinforced my passion for building end-to-end, evaluation-driven systems that move from prototype to deployment — emphasizing reliability and explainability at every step.

Advice for Future GSoC Applicants

If you’re considering applying to GSoC, here’s what worked for me:

1. Start early. Reach out to mentors before the application window. Show genuine interest by engaging with the project’s codebase and community.
2. Write a specific proposal. Don’t be vague. Break your project into milestones with clear deliverables and timelines.
3. Show, don’t tell. If you can submit a small PR or demo before applying, it speaks louder than any resume.
4. Be honest about what you don’t know. Mentors appreciate intellectual humility. They’d rather help you grow than watch you struggle in silence.

GSoC 2025 — Completed ✅

Click above to view the official Google Summer of Code 2025 completion certificate for the project with the Scala Center.

Thank You

A huge thank you to:

• The Scala Center and my mentors for their patience, expertise, and trust.
• Google for running GSoC and creating opportunities for students worldwide.
• The University of Houston and my research advisor for supporting my open-source contributions alongside my academic work.
• The broader Scala community for being welcoming and supportive throughout this journey.
  
  

This blog post marks the completion of my GSoC 2025 project. The code is open-source and available through the Scala Center’s GSoC repository. If you’re a Scala developer interested in RAG, I’d love to hear from you — let’s build the future of intelligent retrieval in Scala together.