Friction Engine: AI Decision Conflict Analyzer

An AI system that reveals hidden value conflicts, tradeoffs, and long-term consequences behind life decisions instead of giving advice.

Comment

Inspiration

Most AI tools today help users decide faster, but not understand deeper.
We noticed that people don’t fail because they lack options — they fail because they cannot see the conflicting values behind their choices.

Friction Engine was built to solve this invisible problem:
turning internal confusion into structured, explainable decision conflicts.

What it does

Friction Engine is an AI-powered decision analysis system that:

  • Detects hidden value conflicts (e.g., Growth vs Stability)
  • Identifies internal contradictions in reasoning
  • Computes a Friction Score (0–100) to measure decision tension
  • Explains why the user is stuck
  • Breaks down root causes of conflict
  • Shows future consequences, regrets, and opportunity costs

It does NOT recommend decisions — it reveals the structure of the dilemma.

How we built it

  • Frontend: React + Vite
  • Backend: Node.js + Express
  • AI Layer: Gemini API (with fallback demo mode)
  • Validation Layer: Structured response schema enforcement
  • UX Layer: Scenario-based decision flows + conflict visualization

The system processes user inputs (options, constraints, values) and transforms them into structured psychological decision maps.

Challenges we faced

  • Designing outputs that avoid generic AI advice
  • Structuring “conflict intelligence” instead of chat responses
  • Ensuring AI responses remain deterministic enough for UI rendering
  • Balancing depth vs clarity in explanations
  • Preventing the system from becoming a “ChatGPT clone”

What we learned

  • The real value of AI is not answers, but clarity of thinking
  • Decision-making is fundamentally a value conflict problem
  • UX matters as much as model quality in AI systems
  • Structured reasoning outputs are more powerful than conversational ones

Future improvements

  • Multi-decision comparison mode
  • Personal decision history memory
  • Group decision friction analysis
  • Mobile-first optimization

Built With

  • express.js-ai:-google-gemini-api-validation:-custom-schema-validation-layer-deployment:-vercel-(frontend)
  • frontend:-react
  • reasoning
  • render-(backend)-other:-friction-scoring-engine
  • scenario-based
  • tailwind-css-backend:-node.js
  • vite
Share this project:

Updates