HomeCache: What If Your Home Could Actually Remember?

Description and use

HomeCache uses a RAG pipeline to store and retrieve information from household documents, enabling natural language queries with source‑aware answers.

Problem

Household information is scattered across invoices, warranties, receipts, and manuals, making it hard to recall where specific details are stored.

Outcome

HomeCache enables quick, trustworthy answers to household queries by leveraging user-uploaded documents and a RAG architecture.

Workflow

• Upload household documents
• Text extraction and chunking
• Generate embeddings for chunks
• Store chunks and embeddings in MongoDB Atlas Vector Search
• Ask a question in natural language
• Convert question to embedding
• Retrieve relevant memory chunks
• Feed question + memories to Gemma-2-2B-IT
• Generate grounded answer with source attribution
• Display answer and supporting document

Evidence

I built HomeCache, an AI-powered memory for your home.

Source

You can upload your household documents, and instead of digging through files one by one, you can simply ask questions about them in natural language.

Source

HomeCache reads and processes it, breaks the information into smaller searchable pieces, and stores those memories in a vector database.

Source

The relevant memories are then passed to Gemma-2-2B-IT, an open-weight model from Google.

Source

HomeCache also keeps track of which documents were used to answer a question.

Source