Team: @Sentinel.AI
- Members: Ankita Shubash Patil, Mithilesh Yogesh Kolhapurkar
- Track: Enterprise Agents
- Course: 5-Day AI Agents Intensive with Google
- Date: November 2025
Sentinel.AI is an AI-powered monitoring agent designed to help developers automatically monitor application performance, detect issues, and simulate ticket/report creation — all powered by a generative AI model (e.g. Google Gemini).
It offers both a web-based UI (via Streamlit) and a command-line interface, enabling seamless integration into different development or deployment workflows.
Key use cases include:
- Monitoring and analyzing log files or application performance metrics.
- Detecting potential issues or anomalies (errors, performance degradation, crashes, etc.).
- Simulating ticket creation for detected issues (e.g., GitHub or Jira), or generating human-readable reports / email summaries.
- Serving as a proof-of-concept for automated monitoring + AI-assisted alerting / reporting pipelines.
- Web UI — A friendly chat-style interface (Streamlit) to interact with the monitoring agent.
- CLI Mode — Run the agent from the command line for integration in scripts or CI/CD pipelines.
- Log & Performance Monitoring — Parse and analyze performance and error logs to detect anomalies or issues.
- Issue Detection & Ticket Simulation — On detecting issues, simulate creating tickets (e.g., GitHub / Jira) or send report emails.
- Mock Log Support — Useful for testing or demonstrations (e.g. with synthetic / mock logs).
- Flexible Configuration — Use a
.envfile to securely provide API keys (e.g.GOOGLE_API_KEY) and configure agent behavior.
- Python 3.x
- (Optional but recommended) Virtual environment
git clone https://github.com/MITHILESHK11/Sentinel.AI.git
cd Sentinel.AI
python -m venv venv # optional
source venv/bin/activate # on Windows: venv\Scripts\activate
pip install -r requirements.txtCreate a .env file in the root directory with your Google API key:
GOOGLE_API_KEY=your_api_key_here
You can copy from .env.example (if present) to get a template.
streamlit run src/app.pyThis will start the Streamlit-based interactive UI where you can view logs, chat with the agent, and get insights on issues.
python src/main.pyThis runs the agent in terminal mode — useful for automation or integration with CI/CD pipelines.
Sentinel.AI/
│
├── src/ # Source code (main logic)
│ ├── app.py # Streamlit web UI entrypoint
│ ├── main.py # CLI entrypoint
│ └── ... # Other modules for log parsing, analysis, reporting, etc.
│
├── prompts/ # Prompt templates used by AI model
├── requirements.txt # Python dependencies
├── .env # Environment variables (e.g. API keys)
├── test_agent_interaction.py # Sample tests / sanity checks
├── test_import.py # Import test
└── README.md # This file
If you are working on small-to-medium scale projects and want automated monitoring + AI-assisted alerting/reporting — without setting up heavy infrastructure — Sentinel.AI can serve as a lightweight yet flexible solution. It can also serve as a base for building more advanced monitoring & observability tools (e.g., real log ingestion, real alerting, integration with real ticketing systems, dashboards, ML-based anomaly detection).
Contributions, suggestions and improvements are welcome! Feel free to open issues or submit pull requests. Some ideas for future enhancements:
- Integrate real logging frameworks / log ingestion (e.g. from production servers).
- Add anomaly detection or ML-based pattern detection on logs/performance metrics.
- Real integration with ticketing systems (e.g. GitHub Issues API, Jira API) for actual ticket creation.
- Email / Slack / Teams / other notification support for alerts.
- Extend UI: richer dashboard, historical logs, trend graphs, filtering, etc.
You can choose an open-source license (e.g. MIT, Apache 2.0, etc.) to specify the terms — (currently no license file is present).
Built and tested using Python + Streamlit + Google Gemini (or any compatible generative-AI model).