Ancient Art Research Assistant

Welcome to the Ancient Art Research Assistant project. This project runs an agent which automates one of my major hobbies: Researching historical art on the internet and blogging about it. This agent searches the web for new images of art and ancient artifacts to add to its collection. It then researches any new finds and produces 100 word blurbs based on the source website.

Live Demo

A live demo of the app can be accessed here.

Installation

Prerequisites

An Apple machine running MacOS the M1/2 chip (not Intel).

Ensure you have Python >=3.10 <=3.13 installed on your system. This project uses Poetry for dependency management and package handling, offering a seamless setup and execution experience.

First, if you haven't already, install Poetry:

pip install poetry

Next, navigate to your project directory and install the dependencies:

1. First lock the dependencies and then install them:

poetry lock

poetry install

Google Cloud Setup

This project runs on Google Cloud and assumes that you have an existing Google Cloud account and some familiarity with Google Cloud services.

Start by installing the Google Cloud client.

brew install google-cloud-cli

Initialize the Google Cloud CLi tool

gcloud init

Create a new Google Cloud Project

GOOGLE_CLOUD_PROJECT="ancient-art-researcher"
gcloud projects create $GOOGLE_CLOUD_PROJECT --name="Ancient Art Researcher"

Ensure that your Project exists

gcloud projects list

Your project should appear in the list

PROJECT_ID                  NAME                        PROJECT_NUMBER
...                         ...                         ...
ancient-art-researcher      Ancient Art Researcher      566563478993

Set your new project as the default

gcloud config set project $GOOGLE_CLOUD_PROJECT

Enable the Could Billing API for your project

gcloud services enable cloudbilling.googleapis.com

Get your billing account and link it with your project

gcloud beta billing accounts list 
gcloud beta billing projects link $GOOGLE_CLOUD_PROJECT --billing-account=[YOUR_BILLING_ACCOUNT_ID]

cd into the finds_viewer folder. This is a web app where you can view all the latest research done by our AI researcher

cd finds_viewer

Enable the artifact repository API for your project

gcloud services enable artifactregistry.googleapis.com

Finds Viewer Setup

Create a .env file and set the following values:

GOOGLE_CLOUD_PROJECT=ancient-art-researcher
GOOGLE_APPLICATION_CREDENTIALS=[path to JSON file with google application credentials]
BUCKET_NAME=[Google Could bucket for storing the images found by the researcher]
FIRESTORE_DATABASE=[Firestore database for saving the research]
FIRESTORE_IMAGE_COLLECTION=[Collection for storing image research findings]
REGION=[Google Cloud region for hosting the web app]

Create the Google storage bucket

gcloud storage buckets create gs//$BUCKET_NAME --location=$REGION

Enable the Firestore API

gcloud services enable firestore.googleapis.com

Create the Firestore database

gcloud firestore databases create --database=$FIRESTORE_DATABASE \
          --location=us-east1

Enable the Google Cloud Run API

gcloud services enable run.googleapis.com

Build and deploy the finds_viewer web application to Google Cloud

./build.sh

Researcher Setup

cd ../researcher

Just like we did for the finds_viewer, create a .env file

OPENAI_API_KEY=[Your OpenAI API Key]
SERPER_API_KEY=[Your Serper API Key]
FIRECRAWL_API_KEY=[Your Firecrawl API Key]
GROQ_API_KEY=[Your Groq API Key]
GOOGLE_APPLICATION_CREDENTIALS=[JSON file with your Google app credentials]
GOOGLE_CLOUD_PROJECT=[Your Google cloud project name]
FIRESTORE_DATABASE=[Firestore database for saving the research]
GCLOUD_IMAGE_BUCKET=[Google Cloud Bucket for storing the images found by the researcher]
OPENAI_MODEL_NAME=gpt-4o-mini

You should be able to run build.sh to build the researcher docker container

./build.sh

To run the researcher locally

./run_docker.sh

Deploying the Researcher to Google Cloud

TODO

Ansible Setup

    brew install ansible
    ansible-galaxy collection install google.cloud

    gcloud iam service-accounts create ansible-service-account --display-name="Ansible Service Account"

    gcloud projects add-iam-policy-binding [YOUR_PROJECT_ID] --member="serviceAccount:ansible-service-account@[YOUR_PROJECT_ID].iam.gserviceaccount.com" --role="roles/compute.admin"

    gcloud iam service-accounts keys create ~/path/to/your-key-file.json --iam-account=ansible-service-account@[YOUR_PROJECT_ID].iam.gserviceaccount.com

    export GCP_AUTH_KIND=serviceaccount
    export GCP_SERVICE_ACCOUNT_FILE=~/path/to/your-key-file.json
    export GCP_PROJECT=[YOUR_PROJECT_ID]

TODO