Who We Are
TSF (Targetetd Seasonal Forecasts Inc.) builds fully automated forecasting technology that gives small and mid-sized merchants the same level of precision used by enterprise analytics teams. We believe that clarity beats complexity — and that forecasting should be transparent, explainable, and practical.
The Science Behind TSF
Traditional forecasting models operate on a single timeline — each day’s value depends on the one before it. That means when reality shifts mid-month, their forecasts can’t keep up; the curve is already locked into yesterday’s assumptions.
The Model of Temporal Inertia (MTI) changes this. TSF forecasts operate on two timelines at once — a sequential timeline for short-term continuity and a seasonal timeline that maps recurring patterns across years. This structure lets each daily forecast value stand independently, so the model can recognize and capture changes within the forecast period instead of missing them.
That independence is what makes TSF capable of producing 30-, 60-, and 90-day forecasts with measured confidence and within-period adaptability — something traditional forecasting simply can’t do.
Each day’s forecast value is selected from 800 independent forecast models — the result of combining 10 seasonal models and 2 seasonal series, each producing 40 different forecasts per combination. The system evaluates each forecast’s historical accuracy and selects the model with the strongest record of performance for that specific date.
The full TSF Engine — the foundation of all products — includes more than 25 seasonal models, each with 3 seasonal series, creating over 3,000 forecast models for monthly forecasts and another 3,000 for quarterly forecasts. Each model is evaluated objectively; none are trained or tuned subjectively.
Learn more about the science behind MTI and TSF’s forecasting process
Data Architecture & Security
Each client’s sales and forecast data is siloed in a dedicated Neon database — there are no shared schemas or connections between clients. Forecast generation occurs within each client’s environment, ensuring complete data isolation and eliminating the risk of cross-connection. Hosting and encryption standards are managed under Neon and AWS SOC-2 / ISO-27001 compliance frameworks.
Founding and Leadership
TSF was founded by Kevin Burk, the creator of the Model of Temporal inertia and architect of the TSF Seasonal Model Library. His research merges scientific rigor with practical implementation, transforming complex temporal theory into tools that business owners can actually use.
Contact
Targeted Seasonal Forecasts Inc.
264 Chapman Rd. Ste. 208 #24609
Newark, Deleware 19702
Want more information?
Lorem ipsum dolor sit amet, consectetur adipiscing elit.