PELICAN

Methodology

Pelican runs three connected systems: idea generation, research, and AI agents. Pick the track you want, then jump to specific sections with the deep links below.

Research

Adaptive, Archetype-Aware Research

Every signal enters a rigorous multi-stage research gauntlet that adapts its analytical lens to the company's specific business model and life cycle stage.

  • Dynamic Classification: Identifying the business archetype to set the correct analytical guardrails, KPI priorities, and failure-mode stress tests.
  • Triple Pillar Underwriting: A forensic audit of management stewardship, competitive moat trajectories, and the tension between market sentiment and institutional flow.
  • Valuation Triangulation: Establishing a qualitatively adjusted Fair Value by cross-referencing five distinct frameworks, including DCF and Net Liquidation Value.
  • Actionable Verdicts: Every report culminates in a definitive rating (Buy, Hold, Sell, or Unclear) paired with probability-weighted expected returns and a "Thesis Killer" monitoring plan.
Explore research methodology →
Idea Generation

Seven Forensic Engines. One Standard Scale.

Pelican's sourcing layer filters behavioral, fundamental, and structural data, distilling millions of data points into a transparent 10-point Pelican Rating.

  • Behavioral Conviction: 13F convergence, Insider buying, and Congress trades.
  • Fundamental & Structural: Special Situations (spinoffs/tenders), Press Release catalysts, Quantitative screens, and Analyst Target dislocations.
  • Pipeline & Performance: High-conviction signals (Rating 8.0+) move directly to our deep-dive research pipeline, with every idea benchmarked and audited to track performance by signal family.
Explore idea generation →
AI Agents

Mixture-of-Experts Research Interface

A specialized conversational layer that converts our proprietary research library and live market feeds into a queryable interface with professional-grade precision.

  • Intelligent Intent Routing: Requests are automatically analyzed and deployed to domain-specific agents specializing in deep filing analysis, quarterly transcript narratives, or real-time market data.
  • Real-Time Data Integration: Bypasses standard LLM "knowledge cutoffs" by querying external APIs to retrieve the latest trading multiples, analyst estimates, and fresh SEC filings as they hit the tape.
  • Auditable Evidence Chain: Every response creates a traceable record with inline citations, linking directly back to the specific report paragraph, regulatory filing section, or market data point used for the answer.
Explore AI agents →