This is a Next.js project bootstrapped with create-next-app.

Step 1. Make sure your node version is 18

node -v

Step 2. Install dependencies:

yarn 

Step 3(Optional). Switching database

Wren-ui uses SQLite as our default database. To use Postgres as the database of wren-ui, you need to set the two environment variable below.

# windows
SET DB_TYPE=pg
SET PG_URL=postgres://user:password@localhost:5432/dbname 

# linux or mac
export DB_TYPE=pg
export PG_URL=postgres://user:password@localhost:5432/dbname

• PG_URL is the connection string of your postgres database.

To switch back to using SQLite, you can reassign the DB_TYPE to sqlite.

# windows
SET DB_TYPE=sqlite
SET SQLITE_FILE={your_sqlite_file_path} # default is ./db.sqlite3

# linux or mac
export DB_TYPE=sqlite
export SQLITE_FILE={your_sqlite_file_path}

Step 4. Run migrations:

yarn migrate
# or
npm run migrate

Step 5. Run the development server:

# Execute this if you start wren-engine and ibis-server via docker
# Linux or MacOS
export OTHER_SERVICE_USING_DOCKER=true
export EXPERIMENTAL_ENGINE_RUST_VERSION=false # set to true if you want to use the experimental Rust version of the Wren Engine
# Windows
SET OTHER_SERVICE_USING_DOCKER=true
SET EXPERIMENTAL_ENGINE_RUST_VERSION=false # set to true if you want to use the experimental Rust version of the Wren Engine

# Run the development server
yarn dev
# or
npm run dev

Open http://localhost:3000 with your browser to see the result.

There are many modules in Wren AI, to develop wren-ui, you can start other modules(services) via docker-compose. In the Start wren-ui from source code section, you've know how to start wren-ui from the source code to develop. To start other modules via docker-compose, you can follow the steps below.

Step 1. Prepare you .env file In the WrenAI/docker folder, you can find the .env.example file. You can copy this file to .env.local file.

# assume current directory is wren-ui
cd ../docker
cp .env.example .env.local

Step 2. Modify your .env.local file You need to fill the LLM_OPENAI_API_KEY, EMBEDDER_OPENAI_API_KEY with your OPENAI api key before starting.

You can also change the WREN_ENGINE_VERSION, WREN_AI_SERVICE_VERSION, IBIS_SERVER_VERSION to the version you want to use.

Step 3. Start the services via docker-compose

# current directory is WrenAI/docker
docker-compose -f docker-compose-dev.yaml --env-file .env.example up

# you can add a -d flag to run the services in the background
docker-compose -f docker-compose-dev.yaml --env-file .env.example up -d
# then stop the services via
docker-compose -f docker-compose-dev.yaml --env-file .env.example down

Step 4. Start wren-ui from source code refer to Start wren-ui from source code section to start wren-ui from source code.

Step 5. (Optional) Develop other modules along with wren-ui

As mentioned above, you can use docker-compose to start other modules. The same applies when developing other modules. From the perspective of wren-ui, if you want to develop other modules at the same time, you can stop the container then spin up the module from the source code.

eg: If you want to develop ai-service module, you can stop the ai-service container then start the ai-service from the source code.

# docker/docker-compose-dev.yaml
wren-engine:
    image: ghcr.io/canner/wren-engine:${WREN_ENGINE_VERSION}
    pull_policy: always
    platform: ${PLATFORM}
    expose:
      - ${WREN_ENGINE_SQL_PORT}
    ports:
      - ${WREN_ENGINE_PORT}:${WREN_ENGINE_PORT}
    volumes:
      - data:/usr/src/app/etc
    networks:
      - wren
    depends_on:
      - bootstrap
    ...
# comment out the ai-service service
wren-ai-service:
    image: ghcr.io/canner/wren-ai-service:${WREN_AI_SERVICE_VERSION}
    pull_policy: always
    platform: ${PLATFORM}
    ports:
      - ${AI_SERVICE_FORWARD_PORT}:${WREN_AI_SERVICE_PORT}
    environment:
      WREN_AI_SERVICE_PORT: ${WREN_AI_SERVICE_PORT}
      WREN_UI_ENDPOINT: ${WREN_UI_ENDPOINT}
      LLM_OPENAI_API_KEY: ${LLM_OPENAI_API_KEY}
      EMBEDDER_OPENAI_API_KEY: ${EMBEDDER_OPENAI_API_KEY}
      LLM_AZURE_OPENAI_API_KEY: ${LLM_AZURE_OPENAI_API_KEY}
      EMBEDDER_AZURE_OPENAI_API_KEY: ${EMBEDDER_AZURE_OPENAI_API_KEY}
      LOGGING_LEVEL: ${AI_SERVICE_LOGGING_LEVEL}
    networks:
      - wren
    depends_on:
      - qdrant
      - wren-engine

ibis-server:
    image: ghcr.io/canner/wren-engine-ibis:${IBIS_SERVER_VERSION}
    ...

Then refer to the README.md or CONTRIBUTION.md file the module for starting the module from the source code.

eg: refer to the ai-service README to start the ai-service from the source code.

We currently do not support multiple projects in Wren AI. You can only have one project at a time. But there is a workaround for this. Since Wren Engine is stateless and we store your semantic model in the database(Sqlite or Postgres), you can switch between projects by switching the database and make sure you deploying after server started.

Tip: Define the DB_TYPE and SQLITE_FILE or PG_URL variable to specify which database you intend to use.

eg:

# start your first project using default database(sqlite by defulat)
yarn migrate
yarn dev

# ... after onboarding and lots of hard work, you want to switch to another project 
# stop the server

# set another sqlite file
export SQLITE_FILE=./new_project.sqlite
yarn migrate
yarn dev

# In the Browser, ... after another onboarding process and hard work
# you can switch back to the first project by setting the first sqlite file
export SQLITE_FILE=./first_project.sqlite

yarn dev  # no need to do migration again

# in the modeling page, click the deploy button to deploy the project to the wren-ai-service.
# your Wren AI is ready to answer your question.

To learn more about Next.js, take a look at the following resources:

• Next.js Documentation - learn about Next.js features and API.
• Learn Next.js - an interactive Next.js tutorial.

You can check out the Next.js GitHub repository - your feedback and contributions are welcome!