Comment on page


Learn more about how Swiss33 is designed
Swiss33 is a freemium algorithmic trading platform that allows users to create their own machine learning models and to deploy them as trading bots on their brokers / exchanges.


Each model has a name, an algorithm, a data source, a granularity, a period, an asset, a list of features (OHLCV / Technical Indicators), a return discrimination, a training set and parameters of the algorithm.
This algorithm stands for "Support Vector Machine". The parameters available are :
  • Kernel
  • C
  • Gamma
This algorithm stands for "eXtreme Gradient Boosting". The parameters available are :
  • Learning rate
Each model has to be trained before being deployed as a trading bot.

Trading bots

Each trading bot has a name, an associated (trained) model, a broker / exchange, a frequency and quantity (of each trade)
  • Robinhood
  • Interactive Brokers
  • Charles Schwab
  • Alpaca Markets
  • TD Ameritrade
  • IG Markets
  • Oanda
  • Degiro
  • TradeRepublic
  • Samco
  • Tradier
  • Trading Technologies
  • Zerodha
  • Coinbase
  • Kraken
  • KuCoin
  • BitStamp
  • OKX
  • BitFinex
  • ByBit
  • Bithumb
  • Gemini
  • Bitget


There is such an abstraction level that you can use a model with a lot of different trading bots ( brokers / exchanges). For instance, you can train a model for BTC/USDT for and deploy with a trading bot connected to, Kraken, Coinbase etc.
The only condition is that the asset of the model must exist within the trading bot's broker / exchange.
For instance, if you train a model for EUR/USD for Interactive Brokers, you can't deploy it on Coinbase, as Coinbase only supports cryptos.
Last modified 7mo ago