Play with the engines. No sign-up, no data required.

Two ways in: visit our playground — a functioning instance of the finsait platform, preloaded with synthetic data — or test the engines with the in-browser illustrations below.

Our playground: the real platform

Want the real thing? Our playground is finsait itself, running on Amadeus Industrial — a synthetic company that receives yesterday's bank feed every night and reprices on real markets. Explore cash position, forecasts, exposures and the full scenario engine. No sign-up required.

Visit the Playground

A living company, refreshed every night

Amadeus Industrial

Yesterday's bank feed, real markets, a new day every morning.

A fictional Iberian industrial group — three entities across Spain, the UK and the US — generated day by day on the real calendar. Collections follow Spanish business sentiment, raw materials reprice on copper, the plant's energy on TTF gas, and the revolver fixes on the real 3-month rate. Every night one more day of transactions lands before the forecast re-runs. Realistic data, an invented company; nothing here is a real client.

Open Amadeus Industrial in the playground
3
Entities in 3 currencies
4
Market drivers in the data
3.5y
Of daily history, growing nightly
60d
Forecast, re-run every morning

Two engines. Open one, play — then feed it your own data.

Each engine opens with a simplified in-browser illustration on synthetic data, for demo purposes. One step further and the actual production engines run on your own CSV — fair use: 3 forecast runs and 10 risk simulations per day, 25 runs lifetime. Hitting those limits? You're ready for the pilot.

Forecasting Engine In-browser illustration

The model learns the company's cash-flow patterns — trend, weekly cycles, month-end spikes — and projects forward with its likely range. You never drag a forecast into the shape you'd like: you move the business drivers underneath it — a currency, demand, rates — and the forecast, its range and the freed-up buffer re-project live. Pick a scenario to set the drivers, then fine-tune them. This is a simplified in-browser model, for demo purposes only.

Group cash position, €M ActualForecastLikely range (8 in 10)

The band is the likely range: 8 times in 10 the cash ends up inside it.

94.2%
Backtested forecast accuracy
8.1%
Forecast variance (25% on spreadsheets)
€22M
Safety buffer released

Synthetic data for illustration. Accuracy and buffer figures are modelled on the fictional dataset, not client results.

Now run the production engine on your own series

First time, we email you a 6-digit code to confirm the address; after that, your email alone signs you back in. No sales follow-up: the address is a gatekeeper, not a lead list — we use it only to enforce fair use and to learn whether engine tests turn into pilots.

Your file is processed in memory to produce the result and is not stored. Runs are counted per verified email and per IP for fair use. Real engines, real math — results depend entirely on your data.

The Cash Advisor Coming soon

We're building a genAI assistant into every finsait deployment — one that understands your data and answers in plain language. Here's the kind of conversation it's being built for.

Amadeus's group cash sits at €163M, two-thirds of it in the Santander pool account in Spain, with roughly $25M across the US accounts earning below-market rates — a candidate for sweeping or short-dated investment.

Scripted sample responses, for illustration only — the Advisor is not yet built. It's on the roadmap; when it ships it will answer from your own connected data.

See it on your own data.

These illustrations run on a fictional company. Visit our playground to see the real platform on Amadeus — or join the pilot and we point the engines at your data.

Join the pilot Visit the Playground