Installation

Get the Respondent community edition running in under 5 minutes

Prerequisites

Verify both are installed:

docker --version
docker compose version

Create the project directory

Create a directory for your Respondent deployment with the required file structure:

mkdir -p respondent/{sources.d,analysis.d}
cd respondent

You will create two files: compose.yaml (tells Docker how to run Respondent) and respondent.yaml (configures the application itself).

Create compose.yaml

Create compose.yaml in the project root:

services:
  respondent:
    image: "docker.io/alevsk/respondent-community:latest"
    command: ["serve", "--config", "/etc/respondent/respondent.community.yaml"]
    restart: unless-stopped
    env_file:
      - path: .env
        required: false
    ports:
      - "${CADDY_HTTP_PORT:-8090}:8090"
    volumes:
      - ./respondent.yaml:/etc/respondent/respondent.community.yaml:ro
      - ./sources.d:/etc/respondent/sources.d:ro
      - ./analysis.d:/etc/respondent/analysis.d:ro
      - respondent_data:/data
    environment:
      - RESPONDENT_DATABASE_PATH=/data/respondent.db

volumes:
  respondent_data:
ⓘ What this does
  • Pulls the community edition image from the Respondent registry
  • Mounts your local config files as read-only volumes
  • Persists the SQLite database in a Docker volume (respondent_data) so data survives container restarts
  • Exposes the UI, API, and WebSocket on port 8090 (override with CADDY_HTTP_PORT env var)

Create respondent.yaml

Create respondent.yaml in the project root with a minimal configuration:

database:
  path: "./respondent.db"

server:
  port: 8090

ingest:
  sources_dir: "./sources.d"

ai:
  enabled: false

logging:
  level: "info"

This is the smallest valid configuration. It uses SQLite for storage, serves the UI on port 8090, and disables AI features. You will expand this file in the Configuration guide.

Start Respondent

Docker Compose is the recommended way to run Respondent — it stitches together the volumes, env file, named data volume, and port mapping for you, and survives reboots via restart: unless-stopped.

docker compose up -d

Alternative: docker run

If you cannot use Docker Compose, the same deployment can be launched directly with docker run. Pass the volumes, env file, and port mapping explicitly:

docker volume create respondent_data

docker run -d \
  --name respondent-community \
  --restart unless-stopped \
  --env-file ./.env \
  -e RESPONDENT_DATABASE_PATH=/data/respondent.db \
  -p 8090:8090 \
  -v "$(pwd)/respondent.yaml:/etc/respondent/respondent.community.yaml:ro" \
  -v "$(pwd)/sources.d:/etc/respondent/sources.d:ro" \
  -v "$(pwd)/analysis.d:/etc/respondent/analysis.d:ro" \
  -v respondent_data:/data \
  docker.io/alevsk/respondent-community:latest \
  serve --config /etc/respondent/respondent.community.yaml

The container exposes the UI, API, and WebSocket on port 8090 — open http://localhost:8090 in your browser once the container is healthy.

Verify the installation

Check that the container is running and healthy:

docker compose logs -f

Look for a log line containing “server listening” – this confirms Respondent has started successfully. Press Ctrl+C to stop following logs.

Open your browser and navigate to:

http://localhost:8090

You should see the Respondent globe interface. The globe will be empty since you have not configured any data sources yet.

✔ Tip

If port 8090 is already in use, set a different port before starting:

CADDY_HTTP_PORT=9090 docker compose up -d

Then access the UI at http://localhost:9090.

Stopping and restarting

# Stop the container
docker compose down

# Start again (data is preserved in the Docker volume)
docker compose up -d

# Restart after a config change
docker compose restart

Next steps

Edit this page on GitHub