Prelio AI -- Champion Risk Scorer

What I Built

Prelio AI scores layoff risk for specific departments at target companies. Sales reps paste a company name, pick the department their champion sits in, and get a 0-100 risk score with a recommended play (Accelerate, Multi-thread, Monitor).

It pulls real layoff history from public datasets, benchmarks against sector competitors, and calculates AI displacement risk per department.

Tools Used

• -Claude Code -- Built the entire app from terminal
• -Next.js 14 -- App Router, React 18, TypeScript
• -Claude API -- Risk analysis engine (claude-sonnet)
• -Polygon.io + Finnhub -- Company and ticker search
• -Google News RSS + GitHub layoffs.fyi -- Real layoff data

How I Vibe Coded It

This started as a sales tool idea: "what if AEs could see layoff risk before it kills their deal?" I described the concept to Claude Code and it built the full stack in one session.

The interesting part was chaining multiple data sources. Claude Code wired up Polygon.io for company search, Finnhub as a fallback, GitHub CSV datasets for historical layoffs, and Google News RSS for recent layoff news. Then it uses the Claude API to synthesize everything into a risk score.

Key Lessons

• -API chaining is where vibe coding gets powerful. One prompt connected 5 different data sources into a single risk score.
• -AI analyzing AI-fetched data. The app uses AI twice: once to fetch and structure data, once to analyze it. Meta.
• -B2B tools are underrated for vibe coding. Everyone builds consumer apps. Internal tools for sales teams have way less competition.