OpenAlpha is being upgraded, and Riverbit’s AI trading experience is taking another step forward.
It is no longer just a chat interface. It is evolving into a true trading agent system that can run continuously, analyze continuously, and execute continuously.
With OpenAlpha, you can now handle two core types of trading workflows:
1. Instruction-based trading
Just tell it exactly what you want:
“Short BTC at 80,000”
“Go long when the 4H EMA21 crosses above EMA55”
“Open a long position now”
“If Trump posts a bullish tweet, go long with 5x leverage using 50% of the portfolio”
2. Strategy-based trading
Turn your trading logic into a prompt, set the execution frequency, and let the Agent monitor the market and act on it continuously. For example:
“Create a 15-minute Chan Theory trading system that goes long at buy signals and shorts at sell signals”
OpenAlpha now begins by identifying user intent, then breaks complex tasks into multiple execution stages:
monitoring price, tracking tweets, analyzing news, classifying signals as bullish or bearish, validating conditions, and finally triggering the trade.
We have significantly strengthened the system in the following areas:
Task-based execution: the Agent is no longer limited to one-off conversations. It can now run as a persistent task.
Event-driven monitoring: supports trigger conditions based on price, tweets, news, funding rates, and more.
Task management: you can see which tasks are currently running at any time and stop them directly.
Prompt transparency: users can view the actual prompts the Agent is using for execution.
Memory: the Agent can retain preferences, rules, experience, and context. Over time, it learns and adapts, becoming a more complete and capable trading operator.
OpenAlpha is evolving from an AI chat entry point into an Alpha Agent that can remember, decompose, monitor, execute, and manage tasks on your behalf.