Myth versus Reality in Options Flow

People see the word algorithm and assume AI. I want to clear that up, because it matters - both for understanding what this platform actually delivers and for appreciating why AI-based alternatives fall well short of what I do here.

The classification engine behind this flow is not AI. It is not a general-purpose model that has been fed data and asked to find patterns. What it is, is thirty-five years of professional experience - encoded as logic, running in real time, applied consistently to every single print the market produces. My experience. My rules. My understanding of how institutional desks build positions, how the CBOE exchange record works, and what distinguishes a genuine directional trade from noise.

If that sounds like a subtle distinction, it is not. The difference between a system built on hard-won expertise and one built on statistical inference is the difference between knowing what you are looking at and guessing.

Why Experience Cannot Be Replaced

The CBOE exchange record is not a clean dataset. It is a continuous stream of individual prints - legs arriving separately, spreads fragmented across timestamps, multi-leg structures that only make sense when you understand how a professional desk actually builds a trade. I have spent my career at the highest levels of buy-side and sell-side business in both London and New York. I ran a major bank’s derivatives and financial engineering group. Reading this tape is second nature to me - not because I learned it from a course, but because I have done it for real, with real capital, in real markets.

Every rule in the classification system comes from that experience. Not from back-testing a statistical model. Not from asking an AI to identify patterns. From understanding, at first principles, what a specific trade construction means - who would do it, why, and what it tells you about their position and intent.

That knowledge cannot be prompted. It cannot be downloaded. And it does not exist in any publicly available dataset that a model could be trained on.

The Testing Never Stops

Building the classification logic was one part of it. The other part is what keeps it sharp: I test the algorithms every day against unusual or ambiguous trades. Not to train a model - there is no model - but to verify that the logic handles new structures correctly, that edge cases are captured, and that the output matches what I would conclude if I were reading that trade by hand.

This is the closest thing to machine learning in what I do - and it is machine learning in the truest sense. Processing power handles the scale. I provide the judgment. The result is that every subscriber gets access to how I see the market, applied to the full tape, without the limitations of human attention span or the inconsistency that comes from watching a screen for six hours straight.

Why Not Just Use CBOE Trade Alert

The standard institutional product is CBOE Trade Alert, at $375 a month. That is the raw data feed - the underlying exchange record, delivered directly. And that is exactly what it is: raw data.

To get anything useful out of it, you first need to learn the proprietary query language well enough to build filters. Those filters are not optional. Without them, the volume of prints is completely unworkable - the noise-to-signal ratio is enormous. But the moment you start filtering, you are making decisions about what to exclude before you have seen what you are excluding. That introduces its own bias, and most people do not have the experience to build those filters well.

And after all of that, multi-leg reconstruction - working out that five separate prints across three timestamps represent one structured spread — is still entirely manual. Doing it accurately, across the full market, throughout the trading day, is simply not something one person can sustain. Things get missed. That is not a criticism; it is just the reality of what that product is.

What I deliver is that problem already solved. The filtering, the reconstruction, the classification - all done before the trade reaches you.

On the CBOE API - AI Is Not the Answer Either

Some will point to the CBOE API at $799 a month as a third route . It is worth being direct about this: the API is pretty much useless for serious real-time flow analysis. Strict data limits and rate constraints make continuous monitoring impractical. You are not getting the tape - you are getting a throttled, partial view of it with significant gaps.

But even setting that aside, the deeper problem remains. An API is a pipe. What comes out of it is still the same fragmented, noisy exchange record. Feeding that into an AI model does not solve anything. A language model has no understanding of how a derivatives desk structures a trade. It has never sat on a floor. It has no concept of how the real money moves versus how it appears to move on the record. It will find statistical patterns in the data - but that is not the same as understanding what the data means. And in options flow, the difference between a pattern and an understanding is the difference between a coincidence and an edge.

AI is only as good as the knowledge that sits behind it. In this domain, that knowledge comes from running a derivatives and financial engineering group at a major bank, from decades on both sides of institutional markets, from years of building and testing a mental model of how large capital actually behaves. That is not something you can replicate with a prompt.

On the Manual Tape Readers

There are people in this space who genuinely can read the tape. A handful are very good. I am not dismissing what they do.

The ceiling is not their ability - it is the process. Manual tape reading is selective by nature. Attention goes where it has been before. Familiar names. Active sectors. The person who caught a clean sweep at 10am is less sharp at 2pm after four hours at the screen - and that is before accounting for the fact that two things can happen simultaneously in different names and one of them will get missed.

I am better at this than any manual reader. Not because I am smarter, but because I am aided by processing power that has no bias, makes no mistakes, and runs from open to close without dropping its attention for a second. Everything is identified, reconstructed, and labelled. Nothing is hand-picked. Nothing is selectively highlighted after the fact. You get the full picture, every day, consistently.

There is nothing else like it in the market.

About the Creator of Rhodie House Options Intelligence:

I spent thirty-five years at the highest seniority of institutional financial markets. My career began in structured products and financial engineering at a major bank, covering the full spectrum of derivative instruments and then structured credit. That work demanded a precise understanding of how complex instruments are constructed, how institutions use them to express views or manage risk, and what a trade structure reveals about the intent behind it.

I subsequently co-founded and built a specialist alternative credit asset management business that grew to over $4 billion in assets under management. I later co-founded and was CEO of a US government securities primary dealer, operating at the highest level of the US fixed income market, which grew to assets in excess of $30 billion.

I am now retired from institutional life and run my own family office, trading equities, fixed income, foreign exchange, and commodities across global markets. Across three decades and multiple institutions, the common thread was always the same: reading positioning, understanding what large pools of capital were doing and why, and knowing what it implied. Options flow is the purest expression of that discipline - institutional intent made visible in real time, for those who know how to read it.

I still maintain the FINRA Series 17, 24, 63, 66 and 99 licenses under the FINRA MQP not because this platform requires them, but because operating to a high regulatory standard is a habit of thirty-five years that I have no intention of abandoning. Options Intelligence is not a brokerage or advisory service. What it is, is a professional-grade information platform run by someone who has always operated within a regulatory framework and understands exactly where the boundaries are.