← Back to companies
[ OK ] Loaded —
[ INFO ]
$ cd
$ ls -lt
01
02
03
04
05
$ ls -lt
01
02
03
04
05
user@intervues:~/$
Design a real-time machine-learning system that predicts next-day closing prices for the S&P 500 constituents using Google search data as the primary signal. Your pipeline must ingest hourly search query logs (≈50 TB/day), filter finance-related queries, engineer lagged sentiment and velocity features, train nightly, and serve daily pre-market predictions with <100 ms latency. The system should backtest on 10 years of data, guarantee no lookahead leakage, and expose an internal dashboard showing Sharpe, drawdown, and live vs. backtest drift.