Finance built with Jev

Work a firm billed over $20,000 for, redone by Jev in 20 seconds for $0.32, and done better according to the author.

Math filters and on-chain rug checks come first, then Jev scores pump potential and guard rails can override it. Every signal is backtested for 24 hours.

Jev classifies 100% of a real tax document corpus at $0.001 per page: 34x cheaper and 6x faster than the LLM pipeline. Open source.

Set a strategy, pick tickers and backtest: trading decisions now run in seconds instead of minutes, about 100x faster and cheaper than an LLM.

Feed it the news and Jev finds the stocks worth buying. Walkthrough in the thread.

Pulls market data, signals, risk and filings for any stock and has Jev call the next 30 days, with financial data from Valyu.

Jev makes the trading calls with a $10,000 budget.

Jev reads the price feed and decides buy or sell, placing real orders on Kuru's on-chain order book every block.

Predict BTC's price one second ahead against Jev, on Base's 200 ms blocks. No sign-up; the account is set up for you.

Jev decides UP or DOWN on Polymarket's 5-minute Bitcoin market; the video shows a profitable call. Open source.

A backtest over all 239 trading days of 2025, based on market data, news, Wikipedia and Google Trends: 8.1M tokens for $0.32.

A 50-day moving-average rule supplies the facts, Jev supplies the judgement, and the code decides the action. Public code.

Paste a token address and Jev re-reads it every 5 seconds: the call, dump risk, who's in control. Each call is graded 15 minutes later.

Before a trade happens, Jev approves the entry, refuses it or waits for one more observation, and can check your own conditions.

Bring your API key or clone it. Its first test trade on 2.4 ETH closed about 0.23% up.