\caption{Illustrative machine-learning, engineering and business plots. (a) Three simulated training-loss curves. (b) Simulated accuracy versus inference latency for the same model names. (c) A simulated first-order sensor response with Gaussian observation noise. (d) Residuals from the same response model, in fixed-width bins. (e) Simulated quarterly revenue by product. (f) Each product's share of total revenue within the quarter. All CSV inputs are original simulated data, not observed measurements or external benchmark results. The sensor noise uses a fixed seed (2026) and standard deviation 0.03. Model and residual columns are supplied in the data; no fit or statistical uncertainty is inferred by Inklet. Data, recipe and figure: MIT, Mark Marosi.}