Chimera Sentinel: AI Threat & Countermeasure Analysis

An AI-powered system that scans biomedical research to identify emerging pathogen threats, suggest countermeasures, and provide supporting scientific references.

Problem Solved

Keeping up with the vast amount of biomedical research to identify potential pandemic threats is a monumental task. Manually searching, reading, and synthesizing information across thousands of papers is slow and risks missing critical connections.

Solution

Chimera Sentinel automates this process using a multi-stage AI pipeline:

1. Knowledge Base: Ingests research papers (PubMed) and genomic sequences (NCBI) into an Elasticsearch index.
2. Threat Analysis: Uses Google Gemini (via Vertex AI) to analyze a user query against relevant documents retrieved from Elasticsearch, generating a structured threat assessment (Summary, Threat Level, Rationale).
3. Countermeasure Suggestion: Extracts keywords from the query, performs a targeted search for relevant vaccine/antiviral research, and uses Gemini to synthesize potential countermeasures.
4. References: Provides citations (PubMed IDs, NCBI Accessions) for the documents used in the analysis.

Key Features

• AI-Generated Threat Analysis Report
• AI-Generated Countermeasure Suggestions
• Supporting References with Clickable Links
• Web-Based Interface (Streamlit)

Tech Stack

• AI: Google Cloud Vertex AI (Gemini 2.5 Pro)
• Search: Elasticsearch
• Backend: Python, Flask
• Frontend: Streamlit
• Data Sources: NCBI E-utils (PubMed, Nucleotide)
• Libraries: Biopython, Pandas, Requests

How to Run

1. Clone the repository: git clone https://github.com/phahim1/project-chimera.git
2. Navigate to the directory: cd project-chimera
3. Install requirements: pip install -r requirements.txt
4. Add your Elastic Cloud Endpoint/API Key and Google Cloud Service Account JSON key path to config.py. Ensure the .json key file is in the project directory (and listed in .gitignore).
5. Run data ingestion (only needs to be done once): python data_ingestion.py
6. Run the backend server: python backend.py
7. In a separate terminal, run the frontend app: streamlit run app.py
8. Open the local URL provided by Streamlit in your browser.

Demo Video

• Watch our short video demonstration showcasing Chimera Sentinel analyzing a potential threat, generating countermeasures, and providing references: https://www.youtube.com/watch?v=c_SwUdnzrBY

Future Work

• Integrate real-time monitoring of news feeds or public health alerts for earlier signal detection.

• Develop more sophisticated visualization for identified threats and countermeasures.

• License

This project is licensed under the MIT License. See the LICENSE file for details.