TalentFlow

TalentFlow is an AI-assisted recruiting platform for small and medium businesses. It helps teams manage jobs, upload resumes, evaluate candidate-job fit, and move applicants through a pipeline from application to interview.

What It Does

• Manage job postings and hiring requirements
• Upload candidate resumes (PDF/DOCX) with server-side parsing
• Use Gemini-powered analysis to extract candidate profile insights
• Score candidate-job match with explainable AI reasoning
• Track applications across pipeline stages (Applied, Screening, Interview, Offer, etc.)
• View interview schedules and hiring metrics in a dashboard

Tech Stack

• Frontend: React, TypeScript, Vite, Tailwind CSS, Radix UI, TanStack Query, Wouter
• Backend: Node.js, Express, TypeScript
• Database: PostgreSQL (Neon) with Drizzle ORM
• AI: Google Gemini (gemini-2.5-flash)
• File processing: Multer, pdf-parse, mammoth

Project Structure

TalentFlow/
  client/            # React frontend
  server/            # Express API and app server
  shared/            # Shared schema/types (Drizzle + Zod)
  uploads/           # Uploaded resume files
  drizzle.config.ts  # Drizzle configuration

Prerequisites

• Node.js 20+
• npm 10+
• PostgreSQL database (Neon recommended)
• Gemini API key

Environment Variables

Create a .env file in the repository root:

DATABASE_URL=postgresql://<user>:<password>@<host>/<db>?sslmode=require
GEMINI_API_KEY=your_gemini_api_key
PORT=5000
NODE_ENV=development

Notes:

• DATABASE_URL is required by both runtime and Drizzle config.
• GEMINI_API_KEY enables resume analysis and match scoring.
• If PORT is not set, the server defaults to 5000.

Getting Started

1. Install dependencies:

npm install

2. Push schema to your database:

npm run db:push

3. Start development server:

npm run dev

App and API are served from the same server process.

Available Scripts

• npm run dev: Start development server (Express + Vite middleware)
• npm run build: Build frontend and bundle backend into dist/
• npm run start: Run production build from dist/index.js
• npm run check: Type-check TypeScript
• npm run db:push: Push Drizzle schema changes to database

Core Routes

Frontend pages:

• / Dashboard
• /jobs Jobs list
• /jobs/new Create job
• /candidates Candidates list
• /candidates/upload Resume upload
• /applications/:id Application details
• /pipeline Kanban pipeline
• /schedule Interview schedule
• /settings Settings

API endpoints (selected):

• GET /api/stats
• GET /api/jobs
• GET /api/jobs/:id
• POST /api/jobs
• GET /api/candidates
• GET /api/candidates/:id
• POST /api/candidates/upload
• GET /api/applications
• GET /api/applications/recent
• GET /api/applications/pipeline
• GET /api/applications/:id
• PATCH /api/applications/:id/status
• GET /api/interviews
• GET /uploads/:filename

Resume Upload Behavior

• Allowed file types: .pdf, .docx
• Max file size: 10 MB
• Uploaded files are stored in uploads/
• Server extracts text and runs AI analysis before creating/updating candidate records

Production Build

npm run build
npm run start

This creates:

• Frontend static assets (Vite output)
• Backend bundle at dist/index.js

Troubleshooting

• DATABASE_URL must be set: ensure .env exists and is loaded.
• AI analysis failures: verify GEMINI_API_KEY and quota; upload still completes with limited fallback data in some cases.
• File upload errors: confirm the file is PDF or DOCX and under 10 MB.

License

MIT