Skip to content

Web Researcher

This Recipe shows an example of a web researcher chatbot.

Core Concepts

Customizing the AgentProcessingUnit
Running the UI

Agents

Conversational Agent

This uses the SimpleAgent template, but needs some customization to enable file uploads and support multiple LLMs.

Speech Agent

This uses the builtin AutonomousSpeechAgent to enable speech-to-text capabilities.

Web Researcher

This is the researcher agent. Customize the prompts, search parameters, etc… to suit your needs.

You will notice that enabled file upload on our AgentProcessingUnit’s primary action.

actions:
- name: "converse"
description: "A copilot that engages with the user."
allow_file_upload: true

We also have a list of available APUs in resources/apus.yaml.

apus:
- apu: MistralSmall
title: Mistral Small
- apu: MistralMedium
title: Mistral Medium
- apu: MistralLarge
...

We did not need to make any customization to support multimedia within the APU, this is turned on by default 🚀.

Try it out!

First let’s fork for Eidolon’s web researcher repository, clone it to your local machine, and start your server.

Terminal window
git clone https://github.com/eidolon-ai/web-researcher.git
cd web-researcher

Then run the server in dev mode, use the following command:

Terminal window
make serve-dev

WARNING: By default the server is running in dev mode which does not persist the machine state between restarts. To start the server in non-dev mode, use the following command:

Terminal window
make serve

This will assume mongodb is running locally on the default port. If you need to connect to a different mongodb instance, you can set the MONGO_CONNECTION_STR and MONGO_DATABASE_NAME environment variables to the appropriate connection string.

The first time you run this command, you may be prompted to enter credentials that the machine needs to run (ie, OpenAI API Key, Google CSE key, and Google CSE Token).

These resources will be saved in the .env file in the project root.

To use other LLM services, you will need to add the appropriate credentials to the .env file in the project root. (ie, ANTHROPIC_API_KEY, MISTRAL_API_KEY, OLLAMA_URL, etc.)

This command will download the dependencies required to run your agent machine and start the Eidolon http server in “dev-mode”.

If the server starts successfully, you should see the following output:

Starting Server...
INFO: Started server process [34623]
INFO: Waiting for application startup.
INFO - Building machine 'local_dev'
...
INFO - Server Started in 1.50s

Next let’s run the ui locally.

Terminal window
docker run -e "EIDOLON_SERVER=http://host.docker.internal:8080" -p 3000:3000 eidolonai/webui:latest

Now Head over to the chatbot ui in your favorite browser and start chatting with your new agent.