Adaptive TDEE Calculator - Research-grounded energy expenditure tracking

I built this adaptive TDEE calculator to solve a specific frustration with mainstream fitness apps: their reliance on static, one-size-fits-all equations and vague activity multipliers. Traditional models force you to pick between four arbitrary checkboxes (like "Lightly Active" or "Moderately Active") which completely ignores how your body actually expends energy. This calculator replaces those generic checkboxes entirely by integrating an AI engine as the frontend input processor. Instead of forcing your lifestyle into a rigid category, the AI analyzes your granular descriptions of daily living, training volume, and lifestyle factors to map out an accurate activity profile. Once the AI defines those precise inputs, they are fed into exact, deterministic mathematical formulas under the hood. The core math shifts the priority to lean body mass rather than just total scale weight, offering a more accurate foundational baseline. From there, the formulas layer on the parsed activity tracking and model dynamic changes like Non-Exercise Activity Thermogenesis (NEAT) and metabolic adaptation over time. It even accounts for less common but highly impactful biological variables, such as the direct caloric and metabolic cost of regular plasma donation. Every single calculation metric, modifier, and formula used under the hood is mapped directly to published, peer-reviewed nutritional science and metabolic research, all of which are cited transparently directly in the web application's footer. The backend is built to be as minimal and fast as possible, using a clean Node.js configuration to serve the single-page application without unnecessary script overhead or tracking bloat. I’m launching this today to get direct feedback from the community on the calculation logic, the usability of the UI layout, and any additional metabolic variables or tracking integrations you think would be valuable to add next.