From LangChain's Research Rabbit repository:
Research Rabbit is a web research and summarization assistant that autonomously goes down the rabbit-hole of any user-defined topic. It uses an LLM to generate a search query based on the user's topic, gets web search results, and uses an LLM to summarize the results. It then uses an LLM to reflect on the summary, examines knowledge gaps, and generates a new search query to fill the gaps. This repeats for a user-defined number of cycles, updating the summary with new information from web search and provided the user a final markdown summary with all sources used. It is configured to run with fully local LLMs (via Ollama)!
You can read more about LangChain's Research Rabbit here!
We'll be using LangChain's Self-Hosted Lite version of LangGraph Platform to deploy our LangGraph application today.
In order to make things slightly easier - we're going to leverage the docker compose method, let's get into what we need to do!
The first thing we'll want to do is as follows:
uv add langgraph-cliNOTE: If you don't have
uvinstalled, you can get it with:pip install uv
Now, we can build the image from our root repository directory with:
langgraph build -t research-rabbitNext, we can create our docker-compose.yaml as follows:
touch docker-compose.yamlEdit the created file with your favourite text editor to reflect the following:
volumes:
langgraph-data:
driver: local
services:
langgraph-redis:
image: redis:6
healthcheck:
test: redis-cli ping
interval: 5s
timeout: 1s
retries: 5
langgraph-postgres:
image: postgres:16
ports:
- "5433:5432"
environment:
POSTGRES_DB: postgres
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
volumes:
- langgraph-data:/var/lib/postgresql/data
healthcheck:
test: pg_isready -U postgres
start_period: 10s
timeout: 1s
retries: 5
interval: 5s
langgraph-api:
image: ${IMAGE_NAME}
ports:
- "8123:8000"
depends_on:
langgraph-redis:
condition: service_healthy
langgraph-postgres:
condition: service_healthy
env_file:
- .env
environment:
REDIS_URI: redis://langgraph-redis:6379
LANGSMITH_API_KEY: ${LANGSMITH_API_KEY}
TAVILY_API_KEY: ${TAVILY_API_KEY}
POSTGRES_URI: postgres://postgres:postgres@langgraph-postgres:5432/postgres?sslmode=disable
Notice that this relies on a .env file. Let's fix that now!
Run the following command to copy the sample .env file.
cp .env.sample .env
Use your favourite text editor to include your own LangSmith API key and Tavily API key!
Now you should have a working instance of your LangGraph application!
You can access your newly deployed system as follows: