Cost Compass

Every AI project is different. Costs vary wildly based on model choice, usage patterns, and architecture decisions. Most teams have no idea what they are going to spend until the bill arrives.

Built a CLI tool that works with AI coding agents. It includes a structured prompt directory that guides agents through cost analysis, model selection, and architecture recommendations. The framework helps you test multiple models and pick the one that delivers 95% of results at 40% of the cost.

• Reduced AI costs by 60% across multiple projects

• Better budget predictions before development starts

• Open source and available on GitHub

Technology Stack