SoonThis project is being finalized. Links and details may change.
Year: 2025ai-ml
RAG About Me
A retrieval-augmented profile that answers scholarship questions with verified sources.
RAGNext.jsLangChainSupabase
Problem
Scholarship reviewers need fast, credible answers to specific questions about my achievements.
Approach
Curated essays, transcripts, and leadership evidence into embeddings; built prompt templates and guardrails to ground responses.
Tools
- Python
- LangChain
- Next.js
- Supabase
- OpenAI API
What I learned
Improved document chunking strategies, metadata design, and evaluation prompts for tone and accuracy.
Next steps
Launch a secure interface with analytics and integrate voice queries during exchange year.