Hey there, thanks for the question. We're AWS (Amazon Web Services) experts by trade, so it's not secret we're going to use their platform to the best of our abilities. We're planning on using their Machine Learning engine to provide alot of the buy and sell signals based on the huge amounts of data we're already ingesting into the platform. You can read a little more about the base capability here: https://aws.amazon.com/machine-learning/details/

From an integration perspective, we're going to be throwing all the price data we have at the ML engine, with the intent to get signals back. This is alot trickier than said however, as there a numerous ways to craft our input into the engine, such as providing technical indicators, order book volume, availability of exchange API during peak times and differences in price between exchanges. On top of those inputs, we must provide many alternative configurations which may alter the results returned, such as technical indicators, the length of time each data point relates too, user defined trading thresholds and a bunch more. Finally, we can use range of prediction models to sample this data through - which initially we'll be focusing on regression models to provide analysis.

Our aim is to backtest these strategies on existing datasets during rises and falls we've already got data on. This will aid aligning the best model we provide to the public after extensive testing.

The analytics provided by this engine will not be available in our free service, and will be for premium users subscribing with FLN tokens.

Please let us know if you've got any other questions!