GoShip
How It WorksChallengesPricingEmployers
Sign InGet Started
0developers
0companies
0assessments scored
How It Works
|Pricing
|Challenges
|Employers
|Privacy
|Terms
GoShip© 2026
Get updates
Twitter·GitHub
Systems nominal
BUILDSENIOR

Build a RAG Chat Interface

120 minutes
0 completed
nextjsairagdocument-processing

Description

Build a RAG Chat Interface

Build a Retrieval-Augmented Generation chat where users upload documents and ask questions with context.

Requirements

  • Document upload: accept PDF, TXT, and MD files
  • Text extraction and chunking (split into ~500 token chunks with overlap)
  • Simple vector search: embed chunks using a basic similarity metric
  • Chat interface with message history
  • Each AI response cites which document chunks were used
  • Highlight relevant chunks in a side panel
  • Clear conversation and re-upload documents

    • Technical Notes

    For embeddings, you can use simple TF-IDF or word overlap similarity. For the AI response, use a mock or real API. Store chunks in memory.

    Bonus Points

  • Real embedding API integration (OpenAI embeddings)
  • Chunk preview with relevance scores
  • Multi-document support with source attribution

    • You will be guided through MCP setup before the timer starts