Interactive Brokers enhances algorithms behind IBot
The algorithms in question concern cases where the robot is not completely sure what a trader is asking.
One of the most apparent pieces of proof of the advancements of artificial intelligence in the world of online trading is IBot, the natural language solution developed by Interactive Brokers Group, Inc (NASDAQ:IBKR).
The solution has enjoyed a series of enhancements in the latest version of the TWS platform and the current efforts of the company aim to further beef up the functionalities of IBot by improving the algorithms behind it. In particular, the company says that in the current (beta) release of TWS, it has improved the algorithms used to generate possible results in cases where IBot is not sure what a trader is asking.
In other words, the latest refinements concern the cases where the bot may not understand a given command and returns a “Potentially misunderstood” warning message, accompanied by a best-efforts result.
This is normal for an AI solution, as such products have to learn and be trained.
IBot already covers a decent variety of commands. These range from “Buy 300 more shares” to “Show my account margin”. The bot also helps traders to also scan the markets for, say, the most active US stock. The solution also provides information on market depth, helps work with option chains, simplifies work with calendar, etc.
Earlier this year, Interactive Brokers integrated its website search with IBot to enable educational and task-based returns. As a result, IBot can answer such queries as “How do I fund my account?” and “How do I transfer funds?” by delivering the top five best information links from Interactive Brokers’ website.
IBot was first rolled out in the fall of 2016 with the key purpose to provide a natural way to trade. Using “natural language” means that traders avoid the disadvantages that come with a standard interface, as they are able to speak (or type) commands without worrying about the precise wording.