AI Realtime Trading Risk Assessor

I want to build an AI-driven module that sits alongside my existing trading setup and delivers continuous, real-time risk assessment for both the forex and cryptocurrency markets. The goal is simple: every open position and pending order should be evaluated second-by-second so I immediately see my true exposure and can take action before a drawdown spirals.

Here’s what matters most to me:
• Real-time risk assessment is the entire focus. Position sizing, margin usage, volatility spikes, correlation shifts, and news shocks all need to feed straight into the model so it can flag elevated risk instantly.
• The first release only has to cover forex and crypto. If your architecture is flexible enough to extend into stocks later, that’s a plus but not required right now.
• Actionable outputs are essential. Whether through a lightweight dashboard, e-mail/SMS alerts, or direct API calls back into my broker, I need clear “risk heat” indicators and recommended de-risking steps (reduce size, tighten stops, hedge, etc.).

I’m comfortable with Python, so a solution built around TensorFlow, PyTorch, or similar libraries will slot easily into my workflow, but I’m open to other stacks if you can justify the benefit. Just make sure any data feeds you rely on (price, order-book depth, news sentiment) are either free or come with clear cost estimates.

Deliverables:
• A working AI model that ingests live forex and crypto data and outputs risk metrics in real time.
• An interface (CLI, small web dashboard, or broker API integration) that surfaces those metrics and alerts.
• Setup documentation so I can deploy, retrain, and maintain the system on my own server.

If you have experience building automated risk engines—or better yet, have already tackled the unique quirks of forex and crypto volatility—I’d love to see your approach. Let’s make staying on the right side of risk a lot easier.