This project implements an orchestrated agent system for processing support tickets using Groq's LLaMA 3.3 70B model. The system combines two specialized agents to retrieve information and process ticket data.
- Python 3.8 or higher
- Groq API key
cd hackerrank-orchestrate
cd codeInstall the required Python packages listed in requirements.txt:
pip install -r requirements.txtThe following packages will be installed:
google-generativeai>=0.5.2- For generative AI capabilitiesrank_bm25>=0.2.2- For BM25 ranking algorithmpython-dotenv>=1.0.1- For environment variable managementpandas>=2.2.0- For data processing
Create a .env file in the code/ directory by copying the provided .env.example:
cp .env.example .envOpen .env and replace ENTER_YOUR_API_KEY_HERE with your Groq API key:
OPENAI_BASE_URL=https://api.groq.com/openai/v1
OPENAI_API_KEY=your_groq_api_key_here
OPENAI_MODEL_NAME=llama-3.3-70b-versatile
From the code/ directory, execute the main script:
python main.py- agent.py - Primary agent for processing and orchestration
- retriever.py - Secondary agent for information retrieval
- main.py - Entry point for the application
- requirements.txt - Project dependencies
- .env.example - Template for environment configuration
Results from the execution are saved to:
support_tickets/output.csv
Review this file to see the processed output from the agents.
The project uses the Groq API with the following configuration:
- Base URL:
https://api.groq.com/openai/v1 - Model:
llama-3.3-70b-versatile - Authentication: API key stored in
.envfile
- Ensure your Groq API key is valid and has sufficient quota
- Verify that all dependencies are installed correctly
- Check that the
.envfile is properly configured in thecode/directory - Confirm that input data exists in the
support_tickets/directory
Today is 15 May 2026; which is the result declaration of this Hackerrank Orchestrate and the below is my result and certification of Excellence
Orchestrate brought together some of the most ambitious builders we've seen - 12,885 registrants from 48 countries, and 24 hours to build a end-to-end AI agent from scratch. Your participation helped make it happen.