The AI Art Diagnostics Agent is a powerful multimodal AI assistant designed to perform in-depth visual analysis and artistic criticism on user-uploaded artwork.
This project utilizes Streamlit for an interactive user interface and the LangGraph (ReAct Agent) framework, powered by the Google Gemini 2.5 Flash model, for advanced image and text processing capabilities.
The agent's primary goal is to provide expert insights into the style, historical period, technique, and potential authentication of an artwork.
- Multimodal Analysis: Capable of processing and analyzing both images (PNG, JPG, JPEG) and textual queries simultaneously.
- Authoritative Art Criticism: Provides detailed analysis covering technique, style, period, and potential artists.
- ReAct Agent Framework: Utilizes LangGraph for structured reasoning and execution flow.
- Interactive UI: Built with Streamlit, featuring chat history, file uploading, and dynamic suggestion chips.
- Bilingual Response: The agent is configured to respond in the same language as the user's last message (Indonesian or English).
| Component | Technology | Description |
|---|---|---|
| Core AI Model | gemini-2.5-flash |
Google AI's multimodal model for image and text analysis. |
| Agent Framework | LangGraph (ReAct) |
For complex and structured agent workflow. |
| User Interface (UI) | Streamlit |
Used to create a fast, interactive web application. |
| Programming Language | Python |
The main development language. |
Follow these steps to run the project on your local machine:
- Clone the Repository:
git clone https://github.com/arifinddc/chatbot-art-diagnostics-agent cd chatbot-art-diagnostics-agent - Install Dependencies:
pip install -r requirements.txt
- Setup API Key: Create a folder named
.streamlitin the project root and add a file namedsecrets.tomlinside it:# .streamlit/secrets.toml google_api_key="YOUR_GEMINI_API_KEY_HERE"
- Run the Application:
streamlit run art_diagnostics_agent.py