PathSmith

forging, hammering out, and structuring your future

Comment

About PathSmith

PathSmith is an AI-powered life decision simulator designed to help users navigate major life choices through cognitive analysis and quantitative modeling. The project combines advanced AI with decision science to detect biases, clarify constraints, and visualize branching future paths before commitment.

Core Purpose

Users input complex life dilemmas (e.g., career transitions, relocation decisions, entrepreneurial ventures) and PathSmith analyzes them across three integrated dimensions:

  1. Bias Detection Engine — Identifies cognitive distortions like loss aversion, sunk cost fallacies, and status quo biases that cloud decision-making
  2. Timeline Forecaster — Projects outcomes across Year 1, Year 3, and Year 5 milestones across multiple competing pathways
  3. What-If Sandbox — Enables interactive branching scenarios with dynamic variable adjustment and real-time metric recalculation

Key Features

  • Cognitive Bias Scanning — AI-powered analysis of psychological fallacies embedded in decision framing
  • Constraint Clarification — Interactive questioning to identify hidden parameters (financial, temporal, relational)
  • Multi-Path Simulation — Quantifies distinct life pathways with scoring across financial stability, personal growth, risk, and time commitment
  • Profile Grounding (RAG) — Users upload resumes or transcripts for personalized context-aware path modeling
  • Interactive Sandbox — Adjustable parameters with live recalculation of 5-year projections and trade-off analysis
  • Configurable LLM Engine — Support for OpenAI, Gemini, or local Ollama models with customizable model selection

Tech Stack

  • Frontend: Next.js 14.2.3, React 18, TypeScript, Tailwind CSS, Framer Motion (animations), Zustand (state management), Axios
  • Backend: FastAPI, LangChain, LLM integrations (OpenAI GPT-4, Gemini, Ollama)
  • Architecture: Full-stack AI application with document processing pipeline and multi-turn decision reasoning

User Experience

The UI features:

  • Interactive console-style interface with decision mode selection (long-term vs. short-term horizons)
  • File upload for RAG grounding (PDF/TXT support)
  • Real-time metric visualizations with progress bars and color-coded path indicators
  • Mock scenario demonstrations showing outcomes for career pivots (startup founder, corporate management, digital nomad)
  • Smooth animations and transitions for progressive disclosure of analysis layers

Philosophy

PathSmith helps users move beyond intuition and bias when facing high-stakes decisions. By combining AI-driven psychological insight with quantitative scenario modeling, it enables users to "forge, hammer out, and structure" their future with greater clarity and confidence.

Built With

Share this project:

Updates