GSoC-2026

[Goldypahal]

Hi everyone!
My name is Goldy Pahal — I'm a second-year B.Tech student specializing in AI & ML at Chandigarh University, India. I found C2SI through the GSoC 2026 organizations list and I'm genuinely excited about the work you're doing in open-source AI tooling.

I'm interested in the PII-Safe — Privacy Middleware for LLM Pipelines project. This project resonates with me because I recently built SHIELD — an AI-powered women's safety system with a 3-layer real-time data pipeline. A core part of SHIELD involved handling sensitive user data (audio + motion) and ensuring it was processed safely before any downstream action. When I read the PII-Safe description, I immediately saw the parallel: intercepting sensitive data before it reaches an LLM is the same architectural problem I've already solved, just in a different context.

My relevant stack includes Python, LangChain, Ollama (local LLM deployment), Firebase, and RAG pipeline design. I'm currently studying MCP (Model Context Protocol) and spaCy NER to make sure I'm fully prepared before submitting my proposal.

I'd love to connect with the mentors and community here. A few things I'm hoping to learn:

• Are there any good first issues I can start contributing to this week?
• Is there a preferred approach for the PII detection layer — spaCy, regex, or a hybrid?

I've already started exploring the c2siorg GitHub repos. Looking forward to contributing to such an impactful project!

Goldy Pahal
GitHub: github.com/Goldypahal
LinkedIn: linkedin.com/in/goldy-pahal-8a6b84332/