Play · Earn · Grow
Your gateway to digital wealth & freedom — a capped, reserve-funded rewards ecosystem built on Polygon.
The Future of Digital Gaming, Rewards & Wealth Creation — Powered by Polygon Blockchain
Connecting to Binance
CGBT AI is a technology-led financial markets company built around a single conviction: the foreign exchange market rewards consistency, and consistency is an engineering problem before it is a trading one.
We build and operate artificial intelligence systems that study currency price behaviour continuously — across every major session, on every major pair — and translate that study into disciplined, rules-bound market activity. No intuition, no fatigue, no drift from the plan.
Around that technology we have built an operation that values transparency over promises, process over personality, and long-term durability over short-term noise. What we publish, we can explain. What we automate, we can audit.
Decisions originate from trained models, not opinions.
Every position sits inside a pre-defined risk envelope.
Activity is logged, reviewable and explainable end to end.
We exist to put disciplined, machine-driven currency market analysis in the hands of everyday participants — with the same rigour, risk controls and transparency that were once reserved for professional trading floors.
We intend to be recognised worldwide as the reference point for artificial intelligence in foreign exchange — a firm judged not by what it claims, but by the clarity of its process, the strength of its safeguards and the trust of the people it serves.
Narrow, testable decisions over broad market calls.
Capital preservation is the first constraint, not an afterthought.
Plain explanations of how the system reaches a decision.
Models are retrained and re-validated as markets change.
We are a team of experienced currency market professionals, quantitative researchers and machine-learning engineers with a combined track record across global financial markets, algorithmic strategy design and risk management.
Our traders bring the market context — session behaviour, liquidity, volatility regimes, the reasons a chart does what it does. Our engineers bring the systems that turn that context into models, and the infrastructure that runs those models reliably at market speed.
That combination is deliberate. A model without market judgement is fragile; market judgement without systems is inconsistent. CGBT AI is built on the belief that neither works alone.
Currency traders, quantitative researchers, ML engineers and risk officers working as one desk.
Every strategy is hypothesised, back-tested against historical regimes and forward-validated before it goes live.
Our data pipeline, model runtime and execution layer are built in-house, not assembled from black boxes.
Segregated duties, change control and continuous monitoring across the whole stack.
Six disciplines that together make up the CGBT AI desk. Each one is owned by named specialists and reviewed on a fixed cycle.
Continuous multi-timeframe study of major, minor and cross currency pairs — structure, momentum, volatility and session behaviour.
Supervised and sequence models trained on years of tick and candle history, re-validated as market regimes shift.
Hypothesis, back-test, walk-forward, live shadow, then deployment. A strategy earns its way onto the desk.
Per-position limits, exposure caps, drawdown circuit-breakers and correlation controls enforced by the system itself.
Low-latency order routing with slippage monitoring, so the trade that reaches the market is the trade the model intended.
Macro calendar tracking, central-bank policy monitoring and cross-market context feeding directly into model inputs.
Four stages run continuously while the market is open. Each stage can veto the next — nothing reaches the market on a single opinion.
Market data streams in from every active session — price, spread, depth and volatility — and is normalised into a single clean series per instrument.
Trained models score the current state of each pair: structure, momentum regime, and whether conditions resemble anything the model has learned to act on.
The risk layer sizes the decision against exposure limits, correlation and remaining drawdown budget — and blocks it outright if any constraint is breached.
Approved decisions route to market automatically, then stay monitored: outcomes are recorded and fed back into the next round of model validation.
Any stage can stop the process. If the risk layer refuses a decision, it is not executed — no override, no exception path, no manual retry around the control.
Currency markets run around the clock across Sydney, Tokyo, London and New York. Our systems are awake for all of it.
Risk rules are code. The system cannot talk itself into an oversized position after a bad run.
Every decision carries the inputs that produced it, so behaviour can be explained and challenged.
Data pipeline, model runtime, risk layer and execution are ours — we can fix and improve any part of it.
Live outcomes flow back into validation. Models that stop earning their place are retired.
Hardened infrastructure, two-factor member access and least-privilege operations throughout.
CGBT is the key to the ecosystem's Global Autopool. It is earned through participation, priced by a published formula with a hard ceiling on growth, and spent inside the platform.
There is no presale and no public offering. CGBT accrues to members through their own participation in the programme — the only way in is to take part.
Price follows an approved formula with a conservative coefficient and a hard ceiling per period, so it cannot be claimed to rise without limit simply because allocations occur.
Autopool entries consume CGBT. Demand for the token comes from something members actually do with it, rather than from expectation alone.
CGBT stays within the platform, which is what keeps its supply, its price formula and its uses under one disciplined set of rules.
The programme is designed around bounds rather than projections. Each of these is enforced in the platform itself, not stated as an intention.
Every position stops at a defined multiple of its own principal. Two limits run at once — one on the daily reward alone, one on everything combined — and reaching either closes the position.
Generation rewards are a bounded share of what a team actually earns, and an unqualified level pays nobody — it does not roll up. Real liability sits below the theoretical maximum by design.
A fixed share of every withdrawal settles immediately on-chain; the remainder becomes a token position. The immediate cash obligation is therefore always a known fraction of any request.
Rank and leadership rewards are released from dedicated reserves after validation. No structural milestone alone obliges a payout that the reserve behind it cannot fund.
The autopool is denominated entirely in CGBT: entries and rewards use the same unit, so a pool can never distribute more than it took in.
Every reward is an append-only ledger entry with a visible source, and every balance is reconstructable from it. Nothing about a member's position is inferred.
Every reward has a visible source · every liability has a ceiling · every position has deterministic accounting.
Each loop feeds the next. Participation drives rewards, rewards build the token, the token powers the pools, and the pools bring participants back to the start.
Package → Active position → Daily reward → Earnings
Team rewards → Qualified generations → Rank progression
Withdrawal → CGBT allocation → Utility → Autopool
Earnings → New position → New lifecycle → More participation
We research, build and operate artificial intelligence systems for the foreign exchange market. Those systems study currency price behaviour continuously and act on it inside a strict, pre-defined risk framework.
Join a platform where market expertise and artificial intelligence work as one system — transparent by design, disciplined by default.