📄 Backgrndy

An AI-powered resume parser and enrichment pipeline that extracts, deduplicates, and structures data into atomic claims using Next.js, Supabase, and Google Gemini.

Backgrndy ingests documents, extracts bullet points, normalizes the data, deduplicates entries using hash-based logic, and leverages AI to enrich and classify atomic claims into a structured dataset.

✨ Features

• Gemini-Powered Parsing: No rule-based section detection. Gemini 3 Flash reads raw resumes and extracts structured JSON for Experience, Skills, Education, and Credentials.
• Normalized Data Model: Companies, entries, and skills are stored in deduplicated tables with full provenance back to source documents.
• Interactive Enrichment Agent: STAR-ordered conversation (Situation, Task, Action, Result) helps produce high-impact resume bullets with quantified metrics.
• Context-Aware Suggestions: The agent uses your existing skills and job titles as silent context for precise rewrites.
• Modular Architecture: Core parsing and staging logic is exposed as a standalone module for integration into other Next.js apps.

🚀 Quick Start

Prerequisites

• Docker and Docker Compose
• Supabase Project (safe alongside other apps via rmc_ table prefix)
• Google Gemini API Key

Setup

1. Clone & Configure

   cp .env.local.example .env.local

   Add your credentials to .env.local:

   • NEXT_PUBLIC_SUPABASE_URL — Your Supabase project URL
   • NEXT_PUBLIC_SUPABASE_ANON_KEY — Supabase anon key
   • SUPABASE_SERVICE_ROLE_KEY — Supabase service role key
   • GEMINI_API_KEY — Google Gemini API key

2. Database Migrations Initialize your Supabase tables. You can use the CLI or run the files in supabase/migrations/ in order via your Supabase SQL Editor:

   npx supabase db push

3. Start Development Server

   make dev

   The application will be available at http://localhost:3000.

🛠️ Docker Commands

We use a Makefile to simplify Docker operations:

• make dev — Start with hot reload for development
• make build — Build production container
• make test — Run unit tests
• make down — Stop and remove containers
• docker compose up --build — Standard Docker command to build and start

💻 Usage

1. Upload: Navigate to /upload and drop your PDF, DOCX, or CSV files.
2. Enrich: Go to /enrich to review the extracted bullet points and trigger AI enrichment via the interactive agent.
3. Library: Browse your full dataset of atomic experience claims and skills in the Library view.

📁 Project Structure

backgrndy/
├── app/                     # Next.js 14 App Router
│   ├── (shell)/             # Main UI pages (Upload, Enrich, Library)
│   └── api/                 # Ingestion & Enrichment endpoints
├── lib/
│   ├── parsers/             # PDF, DOCX, CSV parsers
│   ├── gemini-parse/        # Gemini-specific parsing logic
│   ├── dedup/               # Hash-based dedup logic
│   └── ai/                  # AI provider abstraction
├── components/              # UI components (Shell, Enrich, Upload)
├── supabase/
│   └── migrations/          # SQL database schemas
└── Makefile