Strategy Consulting Agent

OVERVIEW

A strategy consulting agent that breaks business problems into logic trees, formulates testable hypotheses at each node, and stress-tests each hypothesis against targeted analytical frameworks. From the validated structure, it synthesises a strategic recommendation with prioritised next steps.

ARCHITECTURE

The agent runs through a Next.js API route that validates the strategic brief, enforces rate limits, classifies the problem type to construct a hypothesis-driven logic tree, selects applicable frameworks as analytical lenses via OpenRouter, and streams structured findings over SSE. The client renders the decomposition progressively as tree branches with framework-backed evidence and synthesised initiatives and an execution roadmap.

FUNCTIONALITY

• Accepts a business problem, strategic question, or market scenario for analysis
• Builds a hypothesis-driven MECE logic tree before selecting any framework
• Selects frameworks as lenses per branch: P&L decomposition, 3Cs, customer journey, value chain, Five Forces, PESTEL, Ansoff, scenario analysis
• Structured output with branch-level findings and cross-branch synthesis
• Prioritised initiatives with supporting rationale and evidence
• Streaming SSE delivery for progressive rendering of analysis sections
• Per-IP rate limiting with explicit 429 and retry-after contract
• Model fallback chain for availability under provider failures

HOW IT WORKS

The user submits a strategic brief describing a business problem, competitive question, or market scenario. The API validates the input, applies rate limits, then classifies the problem and constructs a hypothesis-driven logic tree with MECE branches. For each branch, the agent selects the most relevant analytical framework, tests the hypothesis with that lens, and streams findings over SSE. The client renders branch-level insights progressively, followed by a cross-branch synthesis that reconciles signals into prioritised initiatives and an execution roadmap.

OUTCOMES

• Accelerates strategic analysis by decomposing problems into testable hypotheses in seconds rather than days
• Improves decision quality through structured hypothesis-driven decomposition rather than generic framework application
• Builds confidence with branch-level evidence that shows exactly which framework validated each conclusion
• Maintains operational safety with rate limits, input validation, and model fallback routing