Uriel olehleh@Uriel-Olehleh

The Hybrid Token-Efficient Routing Agent is a command-line AI system designed to process a collection of natural-language tasks autonomously. Its primary purpose is to reduce remote token usage while maintaining the level of answer accuracy required by the evaluation process. For every task, the system extracts the prompt, identifies its category, estimates its complexity, and selects an appropriate language model. Simple tasks are handled by a bundled local model whenever possible, avoiding remote token costs. Tasks that require stronger mathematical, logical, factual, or programming capabilities are routed to an allowed model through the Fireworks AI service. The system operates inside a Docker container, receives tasks through a structured JSON input file, and records concise answers in a JSON output file. Its routing rules are deterministic and inexpensive, allowing model selection without consuming additional language-model tokens. The application is intended to run without interactive user input and relies on environment variables for model availability, API access, file locations, and local inference settings. This hybrid approach balances answer quality, execution time, and token efficiency.

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.