What is Foresensus?

Are your forecasting activities driving you crazy? Do you know why?

I built Foresensus so you can develop forecasts quickly. With the full power of machine learning, and without having to code, you can create powerful time series forecasts out of the box. While sufficient, I know that anyone looking to develop accurate forecasts, need to also inject circumstance. What do I mean by that? I mean a person knows better than a machine, which events will occur in the future that will impact trends and growth. Luckily, Foresensus allows you to manipulate the forecast to fit for those events and document your rationale along the way.

How does it work?

We leverage facebook's open source project, prophet for our predictions. This procedure was developed to work on forecasting time series data based on an additive model - a regression method where the effects of the individual factors are differentiated and added together.

OK, but explain it like I'm 5...

We are talking about trend lines. Foresensus helps analysts, finance professionals, researchers, developers, and organizational planners build forecasts off time series data. Strictly, that means we need a 1) date and 2) a value such as revenue, units, membership, price, etc.

Once the data is loaded into the tool, a straight-line prediction (linear regression) can be run out of the box for each of your "cohorts" or groups of data you want to predict against. For instance, if you wanted to run a model against Xbox sales for GameStop (cohort 1) vs Xbox sales at Best Buy (cohort 2), you may decide that a linear regression works for cohort 1 but a logistic regression may work best for cohort 2. With Forecast Pro, you can control the regression methods and parameters for each cohort you load into the system.

Once it's modeled, an analyst can then adjust the output line that best matches their expectation - Forecast Pro creates a predicted line (yHat), a less optimistic prediction (yHat lower), a more optimistic prediction (yHat upper), as well as some other basic trends. From there, you can pick the best output line for your cohort and adjust the forecast visually as needed. Manual adjustments to the forecast, give you complete control over the result and often directly reflect your experience and knowledge of upcoming events.

That's great but then what?

Well, when you're done with the modeling, you can export your results in a couple of different formats for use in other systems or, perhaps share the result with your colleagues for further adjustments. You can even compare your results from a previous time period to analyze the effectiveness of your approach last quarter or last year.

This is just the start, and we're excited to build a tool that continues to evolve, especially with new AI capabilities in the market. We do plan on adding additional regressors/models in the future to give everyone quick machine learning capabilities without opening up a code editor.

Try the demo instance -> Demo Login

Sign Up For a Free Trial