Pallas

Pallas is a trading research and analytics platform built to consolidate discretionary and systematic workflows into one instrumented environment. It unifies three core layers. (1) Event intelligence: macroeconomic releases (past/forecast/current) and business news are ingested, classified, and turned into actionable alerts. (2) Execution visibility: broker-aggregated trades and positions are streamed into a high-granularity journal that tracks the full lifecycle (open → manage → close) for analysis beyond end-of-day PnL. (3) Market data: multiple feeds are ingested to compute custom indicators and scanners for opportunity detection. The platform is designed for post-trade attribution that explains *why* PnL happened, not only *how much*. Analysis is structured around timing/price and market-structure diagnostics (e.g., Directional Change-style regimes), and it also hosts experimental uncertainty measures (aleatoric-entropy-like proxies) to contextualize signal reliability and regime noise. Operationally, Pallas emphasizes scale and responsiveness: a local high-precision database, concurrent connectivity to brokers and data vendors, and a UI that stays fluid even under large trade histories.