yura_invrea

For those of you who have questions on what our plugin does, here is a friendly introduction to Bayesian inference and how that fits to Invrea Scenarios: https://invrea.com/blog/bayes_tutorial.php

And that as well :).

To the best of my understanding, they just do _Monte Carlo_ simulation. Our plugin, in addition to just simulation, allow you to do proper statistical inference using Markov Chain Monte Carlo, Sequential Monte Carlo, etc. In other words, our plugin allows you not just define a generative model of some (business) process with some assumptions (guesses), but also to incorporate observations (actuals) you already have.

This is a term used in accounting and so on. Firstly, there is a company budget plan, and then each month (or week, or so on), _actuals_ are being incorporated to the plan.

See, here, e.g., under the Noun section: https://en.wiktionary.org/wiki/actual

Usually we call it an observation, and it is exactly OBSERVE command in probabilistic programming languages like Venture/Anglican, but in our plugin we have decided to try business-(accounting-)related terminology.

feral, thank you very much for your feedback!

>> Just today, making LTV projections, I wanted something like this.

Cool!

>> I have no idea what 'Actuals' means.

This is a way to make an observation. Like OBSERVE in Venture/Anglican (probabilistic programming languages, in Stan it is similar).

We decided to call it ACTUAL, because in accounting and similar fields people often call an observation (known data) an actual.

>> Its frustrating to click a 'Download' CTA and get hit with a webform

If you have already filled the form, I will send you a download link now. If not, please, feel free to fill, and I will send it soon (or tomorrow, we are UK-based).

We just try to track our early users, since it is still very early alpha.

>> "THE SPREADSHEET MODELING TOOL" ...

Good point and useful feedback, thank you! We will discuss it.

Each spreadsheet is a probabilistic generative model, and we allow you to express uncertainty in some cells using functions like GAUSSIAN, NEAR, CHOICE, and also incorporate already known data using function ACTUAL.

Then, the plugin generates valid scenarios given all assumptions and actuals. You can explore scenarios, see averages (expected values), enhance the model, and repeat it until you are satisfied with the prediction quality.

Check out the way state-of-the-art machine learning modeling is possible now in Excel for any spreadsheet. See the example of predicting Rio 2016 gold medal counts.