These community experiments place Jev in a trading system’s decision layer: a program supplies market context and a bounded choice, then separate code handles testing or execution. The collection includes both backtesting and authors’ reports of live trading. Those settings are different, and a demo or a reported trade does not establish a profitable strategy.
New to the model? What is Jev AI explains its state-to-decision interface.
Browse Jev case topics
Separate decisions from execution
Identify whether the source describes a backtest, a simulation, or a live account before interpreting its outcome.
Inspect the state supplied to the model and the permitted outputs; the surrounding system defines the strategy and available actions.
Keep the model’s choice distinct from the execution adapter, position limits, and stop conditions enforced by application code.
What these experiments can demonstrate
A source can show how an API decision connects to a signal pipeline or an order interface. It cannot establish future returns from a short run. JevPlay has not audited the authors’ accounts, backtests, or reported performance. These links document community implementations; they are not trading recommendations or an endorsement of their risk controls.
Selected community demos
Each selection links to its original author’s post. Source text and media come from the existing Jev community archive.
AI Hedge Fund: a backtesting integration
Virat Singh describes adding Jev to the AI Hedge Fund workflow: select a strategy and tickers, then backtest. This post concerns a testing pipeline rather than evidence of live-account returns.
Indra reports an intraday live-account experiment that reached a hard stop after an earlier positive period. The source includes an adverse outcome and a plan to keep adjusting the strategy.
Zade shares a prototype, a demo link, and a repository for local testing with a user-supplied API key. The post reports a single trade; that observation is not a sustained-performance record.
The _trou3 demo illustrates reading structured trading signals and returning decisions. It highlights the input-to-choice interface without providing an independently verified strategy evaluation.
Jarrod Watts supplies an asset pair’s price feed, asks Jev for buy or sell, and connects the choice to Kuru’s order book on Monad. The source demonstrates the execution connection, not guaranteed trading results.
Understand state, allowed outputs, and application actions before interpreting a decision-model integration.
JevPlay is an independent project and is not affiliated with TypeSafe. Community posts, media, and trademarks belong to their original owners. Browse the full Jev case library for more sources.
Virat Singh@virattt
