Skip to content
AZ
Loading
AZ
All Projects
Case Study

AI UX Benchmarking Engine

SaaS platform that transforms weeks of manual UX/CX auditing into minutes - crawls any website, analyses user journeys, and produces AI-generated benchmarking reports with actionable recommendations.

Role
Full Stack Developer + AI Integration
Timeline
2024
Next.jsLangChainLangGraphLangSmithOpenAI APIHasuraGraphQLTypeScriptPython

01The Problem

UX and CX auditing was a weeks-long manual process involving expert reviewers, heuristic checklists, and comparison against industry benchmarks. Small teams couldn't afford it and large teams couldn't scale it.

02The Solution

Built an end-to-end AI auditing pipeline using LangChain and LangGraph with a multi-agent approach: one agent crawls and extracts website content, another runs heuristic UX evaluation, a third benchmarks against industry standards, and a synthesis agent compiles the final scored report. Engineered detailed prompt templates for structured, domain-specific output and built the Next.js report viewer with real-time crawl status updates via GraphQL subscriptions.

03Impact

  • Reduced audit time from weeks to minutes
  • LangGraph multi-agent pipeline with 4 specialised AI agents
  • Structured per-category scores and actionable AI recommendations
  • Real-time crawl status visible to users via GraphQL subscriptions