Hi everyone ![]()
I’m Ashish Mahajan, a GSoC 2025 contributor with NumFOCUS, and I’m very interested in the GSoC 2026 project on a natural language interface for AiiDA using multi-agent AI. I have experience with NLP, deep learning, RAG, and LLM-based agent systems, and I’m currently exploring the AiiDA codebase and documentation.
I wanted to ask for some guidance on how to get involved early in a meaningful way:
-
Is there a Discord (or similar) channel where ideas and early architectural designs for this project are discussed?
-
Are there any existing discussions or design explorations around agent architectures, MCP/tool integration, or NL interfaces that would be useful to follow or contribute to?
I’d be happy to start with design discussions, small exploratory prototypes, or anything else that would be most helpful at this stage. Looking forward to learning and engaging with the community!
Ashish ![]()