Fullstack
YunoDoc | AI-Native Knowledge Management
Transforming document chaos into searchable, AI-driven intelligence using RAG.
Next.jsGenAIRAGPostgreSQLDrizzle ORM
Problem
Manual indexing is prone to human error and makes large document bases practically unsearchable.
Solution
A smart repository that 'reads' and categorizes documents automatically using LLMs and OCR.
Core Stack
Technical Specs
RAG (Retrieval-Augmented Generation) for document-based natural language querying
GEMINI integration for automated multi-label classification
Vector Search implementation for sub-second semantic retrieval
Server-side streaming for efficient handling of large PDF binary files
Gallery
Visual Narrative
Problem Solving
Engineering Complexity
"This section highlights my ability to navigate complex architectural decisions autonomously."
Processing large batches of documents without blocking the main UI thread. Solved using background workers and real-time status updates via Firebase.
70% reduction in document processing time and eliminated manual indexing errors for early testers.
Project Impact & Results