If you've been using Robinflow for any amount of time, you know the problem: thousands of options trades flow through the platform every day. Finding the ones that matter to your strategy means configuring filters for premium, strike distance, volume/OI ratio, trade type, and more.
Today, we're making that dramatically simpler.
What's New
The new AI filter accepts natural language descriptions of the trades you're looking for. Instead of setting 15 individual parameters, you can type something like:
"Show me large bullish sweeps on tech stocks with premiums over $500K expiring within 2 weeks"
The AI interprets your intent, maps it to the appropriate filter parameters, and surfaces matching trades in real time. It works across all three data streams — option flow, dark pool, and signals.
How It Works
Under the hood, the AI filter uses a language model to parse your description into structured filter parameters. Here's what happens when you submit a query:
- Intent parsing — The model identifies which parameters you care about (premium, sentiment, expiry, etc.)
- Value extraction — Natural language values ("large", "tech stocks", "within 2 weeks") are converted to concrete filter ranges
- Filter application — The resulting parameters are applied to the live data stream, same as manual filters
- Reasoning display — You can see exactly how the AI interpreted your query, so you can refine if needed
You can combine AI filters with manual overrides. Set a base filter with natural language, then fine-tune individual parameters if the AI's interpretation doesn't perfectly match your intent.
What's Next
This is just the beginning. We're working on saved AI filter presets, cross-stream filtering (e.g., "show me dark pool prints on tickers with unusual options activity"), and push notifications for AI-filtered alerts.
We'd love your feedback. Drop us a line at support@robinflow.io or reach out on X.
