Five specialized AI agents that produce Wall-Street-grade equity research pitches.
Overview
A parallel multi-agent system that generates equity research stock pitches. An orchestrator acting as a 'Managing Director' resolves dependencies and spawns five agents in phases: business intelligence, SEC financials, and news/sentiment run in parallel, feeding a valuation agent (deterministic DCF + comparable-company models in Python), which is validated by up to 12 deterministic checks before a synthesis agent produces a BUY/HOLD/SELL pitch deck via Gamma. Every data point cites its source; all financial computation is done in Python, never estimated in prose.
Highlights
Stack & tools
Architecture
An orchestrator acting as a Managing Director resolves agent dependencies and spawns work in phases. Three research agents run in parallel on independent data sources, hand off JSON sidecars to a deterministic Python valuation engine, and the run is gated by up to 12 checks before a synthesis agent writes the deck. The design keeps every number auditable: all math is in Python, every claim cites a source.