Hello MIT App Inventor community,
I'm interested in applying for Google Summer of Code 2026 and would like to work on the "Trainable ChatBot interface and AI component" project listed on the ideas page.
I've been following the 2025 GSoC Teachable-LLM project (external trainer + export/import workflow)
Rather than duplicate the 2025 work, I see potential in:
-
Deep App Inventor integration: Moving beyond a external web tool to a first-class platform feature with proper component/blocks integration (similar to PIC/PAC pattern).
-
Productization & UX: Polishing the trainer interface and making the export/import path seamless for learners.
Specific areas I have queries on:
-
Is the 2025 work positioned for continued iteration, or should the 2026 scope be independent?
-
What do you see as the highest-priority gaps between the 2025 prototype and a production-ready learner-facing feature?
-
Any architectural concerns or lessons learned from PIC/PAC that I should keep in mind?
I'd appreciate guidance from mentors and the community on feasibility and scope.
Looking forward to discussing this further!
Best regards,
Prabhas mekala.