01 / CORE ENGINE↗
SETUPCHECK
VOLATILITYTRENDMOMENTUMSIGNAL QUALITY
Find a reason to wait.
Separate potential opportunities from low-conviction conditions using momentum, trend, relative volume, and market structure.
THE NEXT EVOLUTION OF ATOMIC TRADES
Small moves.
Smarter
decisions.
An AI-assisted research lab that combines familiar technical tools with experimental machine learning. Built to find stronger setups, flag weaker ones, and help you understand when not to trade.
✳ Built for clarity, not predictions of guaranteed returns.
A / T INTELLIGENCE CORE 001
QUALITY FILTER ON
SELECTIVE↗
Measure the setup. Question the signal.
ENGINE / 02 ML RESEARCH
Investigate conditions, not certainties.
MARKET ≠ CERTAINTYDATA → CONTEXT → DECISION
001 / BUILT TO QUESTION EVERY SIGNALSCROLL TO EXPLORE ↓
01 / INTRODUCING THE PLATFORMEvery signal deserves
Every signal deserves
a second opinion.
Technical indicators show you what the market is doing. An intelligent research layer helps examine whether the surrounding conditions deserve your attention.
02 / RESEARCH MODEL↗
EXPERIMENTAL / ML
TRAIN → HOLDOUT → EVALUATE
MACHINE LEARNING
Beyond one indicator.
A transparent, short-horizon research classifier examines multiple features together. Always accompanied by validation caveats.
03 / INTELLIGENCE↗
✳
ATOMIC AI
Trend and momentum agree, but weak participation suggests caution.
Explain the market context.↗
AI-POWERED EXPLANATIONS
Know the why.
Optional GPT-powered commentary translates technical conditions into readable analysis. Context, not automated trade instructions.
04 / DEFENSE FIRST↗
RISK PARAMETERSDISCIPLINE / 01
0.50%
EXAMPLE ACCOUNT RISK / TRADE
DEFINED STOPREQUIRED
POSITION SIZINGMEASURED
LEVERAGECONTEXTUALIZED
RISK MANAGEMENT
Risk is not an afterthought.
Work through hypothetical account-risk, stop-distance, and leverage scenarios before considering a position.
02 / THE LIVE RESEARCH LABSee the market.
See the market.
Question the setup.
PUBLIC MARKET DATA
RESEARCH MODE ONLY
RESEARCH MODE ONLY
⌄
⌄
LAST COMPLETED CANDLE—
—
RSI / 14—MOMENTUM
REL. VOLUME / 20—PARTICIPATION
ATR / 14—VOLATILITY
TREND / 21 VS 55—STRUCTURE
Four markets. One view.
Selected chart timeframe · completed Kraken candles · fresh Kijun crosses, trend, VWAP and 1.08× volume. Select a market to explore.
Always watching. All 16 combinations.
SCANNING LIVE CANDLES…Automatic research checks: fresh Kijun cross + EMA trend + VWAP + volume ≥ 1.08×. Both directions, all four timeframes. A 4/4 match shows a notification here. Runs every 90 seconds while this page is visible; email alerts keep their own settings.
YOUR CONDITIONS / YOUR INBOX
Let the setup come to you.
Choose a fresh Kijun-Sen cross and optional confirmations. Research alerts keep running when this page is closed. No orders are placed.
Checking email alert availability…
✳ The lab includes real public OHLC data when available. Offline preview switches to clearly labeled synthetic demo data. Indicators and the ML classifier run locally in your browser. AI explanations require a configured API key.
03 / THE ATOMIC METHODOLOGYHuman logic.
Human logic.
Machine scrutiny.
We start with an understandable system of indicators, then evaluate it against multiple layers of market evidence. No black-box promises. No shortcuts around risk.
01
Detect
Look for trend and structure with EMA, Kijun-Sen, rolling VWAP, momentum, and relative volume.
MARKET FEATURES →02
Challenge
Compare the setup to a small historical dataset with an experimental machine-learning classifier.
RESEARCH MODEL →03
Explain
Use optional AI commentary to translate conflicting signals into plain-language market context.
HUMAN REVIEW →04
Manage
Run risk scenarios. A promising chart never eliminates the possibility of losing capital.
RISK FIRST ✓04 / RISK, MADE VISIBLEMake the
Make the
downside
part of the plan.
Explore how account risk, stop distance, and leverage interact. This calculator illustrates sizing mechanics, not a recommendation for your account.
! Leverage magnifies losses and liquidation risk. The example below excludes trading fees, funding, slippage and gap risk.
POSITION SIZING / SANDBOXHYPOTHETICAL
THEORETICAL POSITION NOTIONAL—
ESTIMATED INITIAL MARGIN—
ILLUSTRATIVE LOSS AT STOP*—
*Before fees, slippage, funding and liquidation.
05 / WHERE IT BEGANBorn from
Born from
the 1% idea.
Atomic Trades began as a trading education concept focused on understanding small cryptocurrency market movements, technical tools, leverage, and discipline.
Founded by Michael Hensel, a crypto market participant since 2017, the new direction asks a more important question: how can machine learning and explainable AI help traders examine the quality of their decisions?
EXPLORE THE RESEARCH LAB ↗06 / QUESTIONS, ANSWEREDLess hype.
Less hype.
More clarity.
Does AtomicTrades AI automatically trade for me?+
No. This research beta does not connect to exchange accounts, access funds, or place trades. It is a market research and education tool.
Is the AI model actually trained on price data?+
Yes. The interactive lab trains a small logistic-regression classifier in your browser using recent historical OHLCV-derived features, then evaluates it on a later holdout segment. This short dataset and simple validation are not sufficient to demonstrate predictive skill or profitable execution.
What does the GPT-powered AI do?+
When configured by the site owner, GPT explains market structure and conflicting indicator readings. The separate statistical model calculates the numerical research score. Language-model commentary is not a verified prediction.
Where does the price data come from?+
Public Kraken OHLC endpoints. The latest unfinished candle is excluded. If the endpoint is inaccessible, the lab uses unmistakably marked synthetic demonstration data instead.
Is there a proven profit track record?+
No verified live performance or profitable trading edge is claimed. Any trading system should undergo extensive out-of-sample testing, cost-aware simulation, forward paper trading, and independent review before real-money use.
STAY IN THE LOOPBuild the edge.
Build the edge.
Not the hype.
Follow the development of the AtomicTrades AI research experience, including model testing, new indicators and feature releases.