Walk-forward tested ML signals for NinjaTrader, TradeStation, MultiCharts. You control the strategy.
Build ensemble models without Python or R. Genetic algorithms find patterns. Walk-forward validated.
Battle-tested across healthcare, manufacturing, energy, finance since mid-90s. General-purpose AI, not market curve-fitting.
ML Signal Generator for System Developers
BioComp Profit generates predictive signals for traders who build their own systems in NinjaTrader, TradeStation, MultiCharts, or custom platforms.
You know basic indicators (RSI, MACD, moving averages) work sometimes but fail in different market conditions. You've considered building ML models yourself but don't want to learn Python, spend months on it, or risk overfitting.
Profit gives you ML-powered signals without the coding barrier.
PROFIT GENERATES SIGNALS. YOU BUILD THE COMPLETE SYSTEM.
Think of it as an advanced indicator—but instead of hand-coding rules, genetic algorithms discover which indicator combinations and model structures actually predict your target. Then you use those signals exactly like you'd use RSI or any other indicator in your strategy logic.
With Profit, you import raw securities data, build indicators then use those indicators as inputs to a modeling process that predicts a "target" (such as the change in the next day's price). The modeling process seeks useful input indicators and performant predictive model types and internal structures. Many models are built and the top performers over the in-sample period are brought into Profit to act as a "committee" of models. The models' signals are combined to create a "System Model" which is more robust. You save your system(s) and each day, or numerous times during the day, update your data and open your Profit systems and view current performance and signals.
Powered by NeuroGenetic Optimizer—battle-tested AI used across healthcare, manufacturing, energy, and finance since the mid-90s. General-purpose modeling, not market-specific curve-fitting.
You can use Profit for end-of-day "Swing" trading, where you take and hold a position (long or short) for some number of days or use the day's forecast to adjust your strategy for day trading, looking to buy dips or short tops depending on the expectation of the day's move direction. Profit also supports reading intra-day data from text files that have date and time in the first column so you can build intra-day models too!
Various features enable you to check the robustness of your systems, including holding back and then processing unseen data and 3D response surface displays that graphically show you models' behavioral complexity.
Indicator sensitivity analysis lists the indicators ranked by predictive value. Learn which indicators are useful for predicting your target variable.
Works with: Stocks, futures, forex, bonds, ETFs (end-of-day or intraday data).
Not suitable for: Options trading.
Preventing Confusion Since 2005
TradeStation = Your workshop
Profit = A specialized power tool in that workshop
You use both together. Profit doesn't replace TradeStation. It gives you better signals to use INSIDE TradeStation (or NinjaTrader, or MultiCharts).
Around 2005, the market started confusing Profit with trading platforms themselves. We're a signal generation tool, not a platform replacement. If you're looking for a complete trading platform, check out NinjaTrader or TradeStation. If you want better ML signals to feed your existing platform, we're here.
Why System Developers Choose Profit
Triple validation (in-sample, out-of-sample, walk-forward) prevents the curve-fitting that kills most backtested strategies. Models are tested on data they've never seen. What works forward, not just backward.
Multiple models vote on each signal—like having 5 expert systems analyzing every setup. When most agree, confidence is high. Genetic algorithm finds which combinations work best together. More robust than single-model approaches.
Point-and-click model building. You don't need Python, R, or programming skills. But if you DO code, Profit supports custom indicators via .NET add-ins. Accessible to non-programmers, extensible for developers.
See which indicators actually matter. Ranked list shows predictive value of each input. Drop weak indicators, keep strong ones, rebuild models. Invaluable for discovering what drives your signals and eliminating noise.
Export signals as simple CSV files. Works with all trading platforms, including NinjaTrader, TradeStation, MultiCharts, ThinkorSwim, Interactive Brokers, MetaTrader, Python, and custom platforms. Super simple integration. Full flexibility.
Genetic algorithms evolve model architectures, not just parameters. Searches billions of possible combinations to find structures that predict your target. Mesh regression captures non-linear patterns basic indicators miss.
Models optimize for profitability, not just prediction accuracy. A 60% accurate model that catches big moves beats a 75% accurate model that predicts noise. Profit targets what matters: your bottom line.
Response surface plots show model complexity visually. See how models behave across input ranges. Identify overfitting before it bites you. Insights no other platform provides.
Real support from people who understand system development. Email, forum, and training classes available. We stand behind our tools. Active user community shares integration examples and strategies.
What we deliver to you
And Why That's Actually a Good Sign
System developers who find profitable edges don't advertise them. Market efficiency is real—if everyone knows about a pattern, it stops working.
We've had customers call to say thanks. One found a lagged correlation between a US security and a French one—watched US close, traded French security, profitable edge. He was thrilled. But he'd never post that publicly.
At our training workshops, traders whisper "come here, look at this" and show us their results privately. They won't share specifics on forums. That's smart trading.
Profit has been used by traders and analysts at Standard & Poor's, Invesco, TransUnion, and other institutional firms—both professionally and for personal trading. We can't name individuals or share their strategies (they wouldn't want us to), but institutional-grade traders choose Profit for signal generation.
What we CAN tell you: Profit has been used by system developers since 2002. The NGO engine behind it has been deployed across healthcare, manufacturing, and energy since the mid-90s. It works. But don't expect traders to prove it publicly—that would defeat the purpose of having an edge.
"The best signal generators don't have flashy testimonials. The traders who benefit stay quiet and keep trading."
What kind of computer do you need?
BioComp Profit has the following technical requirements:
Note: Windows-only. Mac users can run via Parallels, VMware, or Boot Camp.
Stop relying on basic indicators that work sometimes and fail in other market conditions. Generate walk-forward tested ML signals without learning Python or risking months of development time.
Used by system developers since 2002. Powered by NeuroGenetic Optimizer—battle-tested across industries since the mid-90s.
For serious system developers, not tire-kickers.
DISCLAIMER: BioComp Profit is not a trading system. It is a software tool for creating market timing systems. BioComp Profit's output is not, and should not be considered, trading advice. This web page and others associated with it on this web site may make statements regarding performance of market timing systems created using BioComp Profit or for trading and/or trading systems to which BioComp Profit may contribute information. This performance is hypothetical or, such in the case of performance statements by customers or members of the press, cannot be or have not been substantiated by records of actual trading and thus must be treated as hypothetical. Hypothetical or simulated performance results have certain inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not actually been executed, the results may have under- or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity. Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses similar to those shown.