Strategic-Simulation-Agent

Strategic Simulation Agent

A 5-node LangGraph agent that takes a plain-text business decision description and generates probabilistic scenario analysis, quantitative outcome projections, risk factor scoring, and a streamed strategic report.

Architecture

Decision Framer → Scenario Generator → Outcome Simulator → Risk Analyzer → Report Generator

• Node 1 — Decision Framer: Classifies decision type and extracts key parameters
• Node 2 — Scenario Generator: Generates Bull / Base / Bear / Tail Risk scenarios with normalized probabilities
• Node 3 — Outcome Simulator: Projects quantitative outcomes per scenario, computes expected value
• Node 4 — Risk Analyzer: Identifies up to 10 risk factors across 5 categories, computes risk scores
• Node 5 — Report Generator: Streams a comprehensive strategic report, applies rule-based recommendation

Stack

• Backend: FastAPI + LangGraph + Ollama (llama3.2)
• Frontend: Streamlit (3-tab UI)
• LLM: Ollama llama3.2 (local)

Quick Start

Prerequisites

ollama pull llama3.2

Run locally

# Backend
cd backend
pip install -r requirements.txt
uvicorn main:app --reload

# Frontend (new terminal)
cd frontend
pip install -r requirements.txt
streamlit run app.py

Or use Make:

make install-backend install-frontend
make run-backend   # terminal 1
make run-frontend  # terminal 2

Docker

docker compose up --build -d
# Pull model inside ollama container:
docker exec ssa-ollama ollama pull llama3.2

Endpoints

Method	Path	Description
GET	/health	System health check
POST	/api/simulate	Blocking simulation
POST	/api/simulate/stream	SSE streaming simulation
POST	/api/decisions	Store a decision
GET	/api/decisions/{id}	Retrieve stored decision

Key Algorithms

Component	Algorithm
Scenario probabilities	Normalized post-parse: p_i = p_i / sum(all_p)
Expected value	EV = Σ(prob_i × revenue_impact_pct_i) — pure Python
Risk score	risk_score = likelihood × impact — pure Python
Recommendation	Rule-based on EV + overall_risk_score thresholds
Confidence	std(probabilities) < 0.10 → high, > 0.18 → low

Scenarios

Scenario	Description
Bull Case	Optimistic — favorable conditions, strong execution
Base Case	Most likely — moderate conditions
Bear Case	Pessimistic — challenging conditions
Tail Risk	Extreme downside — severe disruption