Agentic Observability

Agentic Observability is a Streamlit-based application that integrates with JIRA and Datadog to fetch data, process it using OpenAI's API(or meta ai API or other), and provide actionable insights via a chatbot interface.

Installation

Step 1: Set Up Ollama (Required for Local LLM Testing)

1. Install Ollama: Follow the Ollama installation guide for Linux. Typically, you can use the following command:

   curl -fsSL https://ollama.ai/install.sh | sh

2. Verify Installation: Check that Ollama is installed correctly:

   ollama --version

3. Pull the phi3 Model: The phi3 model is lightweight and suitable for local testing. Pull it using:

   ollama pull phi3

4. Run Ollama: Start the Ollama service:

   ollama serve

5. Verify Ollama Service: Ensure the service is running on port 11434:

   curl http://localhost:11434

---

Step 2: Test Locally with Fake Data (Recommended for First-Time Setup)

1. Clone the Repository:

   git clone https://github.com/your-username/agentic-ai-observability.git
   cd agentic-ai-observability

2. Set Up a Virtual Environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies:

   pip install --upgrade pip
   pip install -r requirements.txt

4. Set Up Environment Variables: Create a .env file in the root directory and add the following:

   OPENAI_API_KEY=your-openai-api-key
   JIRA_BASE_URL=https://your-jira-instance.atlassian.net
   JIRA_EMAIL=your-email@example.com
   JIRA_API_TOKEN=your-jira-api-token
   DATADOG_API_KEY=your-datadog-api-key
   DATADOG_APP_KEY=your-datadog-app-key
   OLLAMA_MODEL_NAME=phi3
   OLLAMA_HOST=http://localhost:11434

5. Run the Application:

   streamlit run ui/streamlit_chat.py

6. Access the Application: Open your browser and navigate to:

   http://localhost:8502
   

• Prompt test example:

  Show Unauthorized SSH login from datadog.
  

  Show high-priority Jira tickets.
  

---

Step 3: Transition to Real APIs (Optional)

Important Note:
To test with real APIs, you need to comment out the code for fake data and uncomment the code for real API calls in the following files:

• app/jira_service.py
• app/datadog_service.py
• app/ollama_agent.py
• ui/streamlit_chat.py

Once you have tested the application with fake data, you can configure it to use real APIs for JIRA and Datadog.

1. Update the .env File: Ensure your .env file contains valid API keys and credentials for JIRA and Datadog.

2. Run the Application:

   streamlit run ui/streamlit_chat.py

3. Verify the APIs:

   • Ensure that your JIRA and Datadog credentials are valid.
   • Check that the application fetches real data from the APIs.

---

Docker/Docker Compose Setup

1. Ensure Docker and Docker Compose Are Installed:

   • Install Docker.
   • Install Docker Compose.

2. Clone the Repository:

   git clone https://github.com/your-username/agentic-ai-observability.git
   cd agentic-observability

3. Set Up the .env File: Create a .env file in the root directory with the following:

   OPENAI_API_KEY=your-openai-api-key
   JIRA_BASE_URL=https://your-jira-instance.atlassian.net
   JIRA_EMAIL=your-email@example.com
   JIRA_API_TOKEN=your-jira-api-token
   DATADOG_API_KEY=your-datadog-api-key
   DATADOG_APP_KEY=your-datadog-app-key
   OLLAMA_MODEL_NAME=phi3
   OLLAMA_HOST=http://ollama:11434

4. Build and Start the Services:

   docker-compose up --build

5. Verify the Services:

   • Check that the chat-ui service is running on port 8501.
   • Check that the ollama service is running on port 11434.

6. Access the Application: Open your browser and navigate to:

   http://localhost:8502
   

---

Troubleshooting

Common Issues

1. Error: Port Already in Use:

   • Stop any process using the conflicting port:

     netstat -ano | findstr :8501  # Replace 8501 with the conflicting port
     taskkill /PID <PID> /F       # Replace <PID> with the process ID

     Another solution is to change the mapped port in the docker-compose.yml file.

2. Error: Failed to Connect to Ollama:

   • Ensure the ollama service is running:

     docker-compose up -d ollama

   • Verify the phi3 model is available:

     docker exec -it <ollama-container-id> ollama list

     If not, download it:

     docker exec -it <ollama-container-id> ollama pull phi3

3. Error: File Not Found:

   • Ensure the data directory and its files (sample_jira_tickets.json, sample_datadog_logs.json) are present in the project.

---

Author

• Jodionísio Muachifi - GitHub Profile

License

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