Local Weaviate Semantic Search PoC

This project demonstrates how to run Weaviate locally and use it for semantic search over JSON documents.

Scope

• Run Weaviate locally with Docker Compose
• Create a Weaviate collection/schema
• Load JSON data into Weaviate
• Generate embeddings locally with SentenceTransformers
• Search documents by semantic similarity
• Verify that data is indexed and searchable

Requirements

• Docker
• Docker Compose
• Python 3.10+

Run locally

Start Weaviate:

docker compose up -d

Create the schema:

python scripts/01_create_schema.py

Import the sample documents:

python scripts/02_import_data.py

Search the data:

python scripts/03_search.py

Verify that objects are stored in Weaviate:

python scripts/04_check_data.py

Project structure

weaviate-research/
├── docker-compose.yml
├── requirements.txt
├── data/
│   └── documents.json
├── scripts/
│   ├── 01_create_schema.py
│   ├── 02_import_data.py
│   ├── 03_search.py
│   └── 04_check_data.py
└── README.md