@larry063863

This AI agent is built for Track 1 of the AMD Developer Hackathon. Its primary goal is to act as a smart router for various natural language processing tasks. The agent handles eight specific capability categories, including factual knowledge, mathematical reasoning, and code generation. To achieve efficiency, the agent uses a routing strategy. It runs smaller local models for simple tasks because these local models cost zero tokens toward the final score. It only calls the premium Fireworks API for complex tasks that require higher reasoning capabilities. This approach allows the agent to control AI spending without sacrificing user experience. The agent follows all strict hackathon requirements. It reads tasks from /input/tasks.json and writes results to /output/results.json. It does not hardcode API keys or model IDs. Instead, it reads all configuration values from the environment variables injected by the grading harness. All external calls are routed through the required FIREWORKS_BASE_URL to ensure accurate token tracking. The agent is containerized as a Docker image and built for the linux/amd64 architecture to ensure it runs correctly on the evaluation system. It is designed to start quickly and stay within the memory limits of the grading environment. By prioritizing accuracy and intelligent routing, this agent provides a reliable solution for managing AI workloads effectively.

To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. ADP members can access AI tutorials, experts, and community support. In addition, new members can claim a free month of DeepLearning.ai pro, $100 in AMD GPU cloud credits, and $50 in Fireworks AI API credits. All hackathon participants will receive on hackathon launch additional compute and API credits for their projects. Ready to build what's next on AMD? Sign up with AMD below to secure your spot.