RoleCall.ai

Land the job you need. RoleCall fills your pipeline with real opportunities from Australian job boards, triages them with AI, and helps you apply at scale — so you can focus on landing the right role, not drowning in tabs.

Built for real people getting back into the workforce, changing careers, or just tired of the soul-crushing manual search.

---

What it does

1. Scrapes job boards automatically — Indeed AU, SEEK, Jora, SmartJobs QLD, EthicalJobs, Sunshine Coast Council Careers. Runs on a schedule you set (default: every 2 days).
2. AI triage — Every new listing gets scored against your profile. Good fits go straight to "Recommended." Not-quite-right jobs land in "Not Right Now" with an explanation of why.
3. Kanban pipeline — Track jobs from Recommended → Applied → Interview → Offer. Drag and drop on mobile or desktop.
4. Multi-user — Each person gets their own search profiles, preferences, and pipeline. Built for households, not just individuals.
5. Self-hosted — Your data, your instance. Deploy on Railway, Fly.io, or any Docker host in minutes.

Tech Stack

Layer	Tech
Frontend	Next.js 15 (App Router), React, TypeScript
Styling	Tailwind CSS v4, shadcn/ui
Auth	Better Auth (email/password)
Database	PostgreSQL 16 + Drizzle ORM
Job Queue	BullMQ + Redis
Scraping	Crawlee (Playwright + Cheerio)
AI	OpenAI GPT-4o-mini via Vercel AI SDK
Drag & Drop	@dnd-kit
Monorepo	pnpm workspaces + Turborepo

Getting Started

Prerequisites

• Node.js 20+
• pnpm 10+
• PostgreSQL 16
• Redis 7
• OpenAI API key (for AI triage)
• Webshare account (for residential proxies — needed for Indeed/SEEK/Jora)

Quick Start (Docker)

git clone https://github.com/your-username/rolecall.git
cd rolecall
cp .env.example .env
# Fill in your environment variables
docker-compose up

Open http://localhost:3000, create an account, set up a search profile, and you're off.

Local Development

pnpm install
cp .env.example .env
# Fill in your environment variables

# Start PostgreSQL and Redis (via Docker or locally)
docker-compose up postgres redis -d

# Run database migrations
pnpm db:migrate

# Start everything
pnpm dev

This starts the Next.js web app on :3000 and the background worker simultaneously via Turborepo.

Deploy to Railway

1. Create a new Railway project
2. Add PostgreSQL and Redis services from the Railway template library
3. Add two services from this repo:
   • web → uses apps/web/Dockerfile
   • worker → uses apps/worker/Dockerfile
4. Set environment variables (see below)
5. Deploy

Railway configs are included at apps/web/railway.json and apps/worker/railway.json.

Environment Variables

Variable	Required	Description
DATABASE_URL	Yes	PostgreSQL connection string
REDIS_URL	Yes	Redis connection string
BETTER_AUTH_SECRET	Yes	Auth encryption key (openssl rand -hex 32)
BETTER_AUTH_URL	Yes	Public URL of the app
OPENAI_API_KEY	Yes	For AI job triage
WEBSHARE_PROXY_HOST	For scraping	Webshare residential proxy host
WEBSHARE_PROXY_PORT	For scraping	Webshare proxy port
WEBSHARE_PROXY_USER	For scraping	Webshare username
WEBSHARE_PROXY_PASS	For scraping	Webshare password

Project Structure

apps/
  web/          → Next.js 15 frontend + API routes + server actions
  worker/       → BullMQ consumer + Crawlee scrapers + AI triage
packages/
  database/     → Drizzle schema, migrations, query functions
  shared/       → Types, constants, utilities shared between apps

How the Pipeline Works

Job boards ──► Scraper (Crawlee) ──► Dedup ──► AI Triage ──► Pipeline
                                                  │
                                         Score + reasoning
                                                  │
                                    ┌─────────────┴─────────────┐
                                    ▼                           ▼
                              Recommended                  Not Right Now
                              (score 70+)                  (with AI explanation)
                                    │
                                    ▼
                        Applied → Interview → Offer

Each search profile defines:

• Keywords — what roles to search for
• Location + radius — where to look
• Boards — which job sites to scrape
• Qualifications & preferences — used by AI to score fit
• Schedule — how often to check for new listings

Scraper Status

Board	Status	Method
SmartJobs QLD	Working	Cheerio (no proxy needed)
Indeed AU	Working	Playwright + residential proxy
SEEK	Stub (needs tuning)	Playwright + AU residential proxy
Jora	Stub (needs tuning)	Playwright + residential proxy
EthicalJobs	Stub (needs tuning)	Cheerio
SCC Careers	Stub (needs tuning)	Cheerio
LinkedIn	Skipped	Enterprise-grade anti-bot

SmartJobs and Indeed are fully implemented. The remaining scrapers have the structure in place but their CSS selectors will need adjusting against live sites — job board markup changes frequently.

Contributing

Contributions welcome. The biggest areas that need help:

• Scraper selectors — SEEK, Jora, EthicalJobs, and SCC Careers stubs need real-world selector validation
• More job boards — PRs adding new Australian (or international) board scrapers are very welcome
• Tests — Unit tests for parsers, dedup logic, and Zod schemas
• Accessibility — Screen reader support, keyboard navigation improvements

Please open an issue before starting large features so we can discuss approach.

License

AGPL-3.0 — You're free to use, modify, and self-host RoleCall. If you deploy a modified version as a service, you must open-source your changes under the same license.

For commercial licensing inquiries, reach out via the issue tracker.