The Suretiform AI engine, explained without the mystique
Artificial intelligence is the least understood and most overclaimed technology in finance, so this page does the opposite of marketing: it explains precisely what the engine does, what data goes in, what comes out, where its limits are, and why it supports your decisions rather than replacing you. If you finish this page thinking "so it is careful statistics, run at machine speed, with honest edges", we have done our job.
The one-sentence version: the engine continuously processes market data from shares, currencies, gold and digital assets, converts it into structured signals, and executes strategies within limits you set, while you keep every decision that matters.
Registration is open. Places are limited.
By creating an account, you agree to our Privacy Policy · Terms of Use.
A personal manager will contact you within 24 hours.
What the AI actually is
Its role, and the role it deliberately does not take.
Technically, the engine is a family of statistical models, including machine-learning classifiers trained on historical market behaviour, running continuously across live data. Its job is pattern recognition at scale: detecting recurring relationships between price, volume, volatility and time that recur often enough to trade, and expressing each finding as a probability with a confidence level, never as a certainty. Strategies then act on the probabilities that clear their thresholds and ignore the rest.
What it is not: a mind, a guru, or a replacement for your judgement. It does not "understand" the news the way you do, hold opinions, or get lucky hunches. It is a support instrument, closer to a tireless analyst with perfect recall and no instincts at all than to the science-fiction version. We are blunt about this because the industry's habit of overclaiming is precisely what makes people either fear the technology or trust it too much, and both mistakes cost money.
How the technology works, stage by stage
Data collection
Prices, volumes, volatility measures and market-structure data stream in continuously from venues and providers, validated for gaps and anomalies on arrival.
Intelligent processing
Models clean, normalise and cross-examine the stream: what is noise, what is signal, what relationships are holding today and which are drifting.
Continuous monitoring
Every market on your watchlist is evaluated against every active strategy's conditions, around the clock, weekends included, with no fatigue and no favourites.
Structured presentation
Findings surface as signals with confidence levels, executed to your guardrails and reported in plain language, so what happened is always explainable.
What the engine analyses
Price movement
Direction, speed and structure of price change across timeframes from minutes to months: trends, ranges, breakouts and exhaustion patterns, measured rather than eyeballed.
Trading volumes
Activity levels and surges that confirm or contradict price moves, including volume that arrives without price follow-through, often the most informative kind.
Volatility
How much markets are moving right now versus their recent norms, which drives position sizing and triggers the volatility pause when conditions exceed your band.
Trends and history
Long-run dynamics and how current conditions compare with thousands of analogous historical windows, the raw material for probability estimates.
Cross-market relationships
Correlations between shares, currencies, gold and digital assets, so the system knows when two positions are effectively one bet and when genuine diversification exists.
Changing conditions
Regime shifts, when the market's character changes character rather than merely price: the engine flags these, because strategies tuned to one regime often misbehave in another.
What AI genuinely adds
Speed and breadth
The engine evaluates every instrument on your watchlist every few seconds, across four time zones and two asset classes, and reacts in milliseconds when a threshold clears. No human reads that much, that fast, for that long.
Constant monitoring
It does not sleep, take holidays or get bored on a quiet Tuesday. Continuous markets get continuous attention, which is the entire premise of automation done properly.
Time returned to you
Members typically spend minutes a week on oversight rather than hours a day on watching. The time you save is the quiet, unadvertised benefit almost every member names first.
Discipline, not mood
Rules apply identically on day one and day four hundred, in drawdowns and rallies alike. The engine cannot panic, and that alone removes a whole category of human error.
Usable by anyone
Beginners get institutional-grade process without needing to understand it; experienced members get a tireless analyst they can supervise. Both get reporting that assumes you are intelligent but busy.
Who the AI direction suits
People short on time
If work, family or travel makes daily market attention impossible, this is the core use case: full-time process, part-time oversight, results you can audit after the fact.
People who want analysis without doing it
You do not need to enjoy charts to invest sensibly. The engine and your manager carry the technical load; you carry the decisions that require your goals and your risk tolerance.
Sceptics, especially
If you distrust AI claims, good; so do we. This page exists so you can inspect the machinery rather than take it on faith, and your manager will happily explain any part of it, including the parts that are just careful statistics and hard work.
Using the AI, in four steps
Register
Create the free account; verification usually clears within one business day.
Activate
Choose markets, strategy profile and guardrails with your manager, then fund from NZ$400.
Meet the tools
Explore the dashboard: signals, positions, thresholds and the monthly report format.
Supervise
Check alerts, read reports, adjust with your manager, and stay the decision-maker throughout.
A worked example of a reaction
A concrete scenario, simplified but faithful to how the system behaves. Suppose a strategy tracks a basket of large-cap shares, and one afternoon a major index gaps up 1.8% on a policy announcement. The engine immediately notes the volume surge, checks whether the move cleared its volatility band, and compares the setup against historical analogues in its training window. Suppose the pattern matches a class that historically continued for several sessions about 61% of the time. The strategy's continuation rule requires 60%, so the signal clears, barely.
What happens next is the part that matters: the engine sizes the position according to the volatility now present, not the volatility of a quiet morning, opens it inside your exposure cap, and registers the exit conditions, a give-back threshold and a time stop, at entry. If the following session brings a sharp reversal instead, the exit rule fires without hesitation or hope, the loss is the sized one you agreed to in advance, and the monthly report explains the whole episode in three plain sentences. That is the entire magic: probabilities respected, sizes disciplined, outcomes recorded. The engine may still have been wrong; it simply refuses to compound being wrong with being undisciplined.
Questions about the AI
Meet the engine with nothing at stake
Register free, ask your manager every sceptical question you have, and only then decide whether to fund anything. Understanding the tool costs nothing; using it starts at NZ$400 with no monthly fee. Markets carry risk regardless of who, or what, does the analysis.