I built a Resume Killer using Next.js, Neo4j, and Gemini

Source: Dev.to

The Problem

I was tired of ATS bots rejecting me. I knew I had the skills, but my resume didn’t show it.

The Idea

What if I could visualize my GitHub commit history as a dependency tree?

The Stack

Frontend

• Next.js 14 (App Router)

Database

• Neo4j – a graph DB that’s perfect for skill trees

AI

• Gemini – to analyse “Impact” vs “Churn” in code

How It Works

• It fetches your public repositories.
• It maps languages to nodes (e.g., React → Frontend).
• It calculates a “Market Readiness” score.

Try It Out

I just launched the MVP today. It’s free for devs. I’d love to see if the graph accurately reflects your stack.

👉

The “Magic” Query

Here’s the actual Cypher query I use in Neo4j to map the skill dependencies. It finds a user, matches their commits, and calculates the weight of the relationship:

MATCH (u:User {username: $username})-[r:COMMITTED]->(c:Commit)
WHERE c.impact_score > 50
MATCH (c)-[:USES_LANGUAGE]->(l:Language)
MERGE (u)-[s:HAS_SKILL]->(l)
ON CREATE SET s.weight = 1
ON MATCH SET s.weight = s.weight + 1
RETURN u, s, l

(This is a simplified version, but it shows how skills are weighted based on actual usage rather than just claimed.)