Jev Financial Trading Demos

Routing wallet decisions from market signals, fees, and model cost during a 22-hour trading run

Classifying observed trades from wallet history and graph context while keeping expensive models out of routine decisions

Integrating typed market decisions into a live trading setup and completing its first trade

Opening a trading risk gate only when market scoring agrees with a reconstructed network of 166,700 fruit-fly neurons

Combining tracked-wallet activity, entry assessment, and Astra explanations to manage positions autonomously over 17 hours

Betting on Bitcoin's next-second direction against typed decisions while testing pace with 200 ms blocks

Evaluating Binance BTC order-book data to paper trade a 1,000 USDT position in five-minute prediction markets

Applying a no-look-ahead momentum strategy to grow a simulated $100,000 stock portfolio to nearly $210,000

Classifying 30 synthetic Arabic and English market-news cases across signal, direction, and event type

Sorting wallet charges, weighing urgency, and feeding bills or funds in amounts from one cent to $100

Letting Grok and Muse provide context, making SKIP, YES, or NO decisions in about 80 ms, and filling Polymarket orders within budget

Scoring 536 wallets and making 52,110 memecoin-trading decisions in 21.6 seconds for $0.37

Reading each Robinhood wallet move, scoring it before copying, and sizing entries inside a Grok-built trading harness

Passing 312 newly launched memecoins through fixed gates, rejecting all but three candidates in one night

Retuning an existing seven-agent pipeline for NFT mints by changing four scoring questions without new code

Gating live trading tape with GROKBOT, Hermes, and typed decisions across 96 closed trades and a reported +30R

Switching a WMON/USDC agent from every-block trading to tape following and range fading through one plain-English instruction

Scanning live markets, tracking sentiment, filtering setups, and executing trades during an autonomous $500 run

Evaluating the five-minute BTC UP/DOWN market and placing a prediction order through an open-source Polymarket agent

Estimating edges for every Sunday NFL game and selecting five reported bets from fresh search context

Making 1,260 buy, sell, or hold decisions for $0.11 and reporting that every strategy underperformed doing nothing

Running an autonomous $50 trading bot and reporting a balance increase to $8,310 without manual intervention

Coordinating ten automated traders across multiple desks and one shared portfolio with live risk and P&L tracking

Watching onchain and offchain data in real time and reporting 90 autonomous fills with a 74.6% hit rate