Skip to content

setup.md

Latest commit

 

History

History
49 lines (33 loc) · 1.13 KB

setup.md

File metadata and controls

49 lines (33 loc) · 1.13 KB
Raw

Setup

#setup

Prerequisites

  • Python 3.10+ (3.12 recommended)
  • An OpenAI API key, or AWS credentials with Bedrock access (for generation)
  • Optional: a HuggingFace API token, only if using hosted embeddings

1. Install

python -m venv .venv
source .venv/bin/activate
make install

2. Configure

cp .env.example .env

Set JWT_SECRET to a real random value (don't ship the default), and either OPENAI_API_KEY or Bedrock credentials. Choose your vector store with VECTOR_STORE=faiss or VECTOR_STORE=chroma (see the README's comparison table if you're unsure).

3. Run

make run

4. Verify

curl http://localhost:8000/ping
# {"status": "OK"}

Switching vector stores

Change VECTOR_STORE in .env and restart — no code or data migration between stores is provided, since they use different on-disk formats. If you switch after already ingesting documents under the old store, re-ingest them under the new one.

Docker

docker build -t docintel-intermediate .
docker run --env-file .env -p 8000:8000 -v $(pwd)/data:/app/src/data docintel-intermediate