LeadFlow — Lead Generation & Email Enrichment Platform

A unified platform for three lead-scraping pipelines with email enrichment via Anymailfinder.

---

Tech Stack

• Next.js 16 (App Router) + TypeScript
• SQLite via Prisma 7 + better-sqlite3
• shadcn/ui + Tailwind CSS
• Anymailfinder API v5.1 — email enrichment (bulk JSON + individual search)
• Vayne API — LinkedIn Sales Navigator scraping
• Apify — Google SERP scraping

---

Setup

1. Install dependencies

cd leadflow
npm install

2. Configure environment

Copy .env.local.example to .env.local:

cp .env.local.example .env.local

Edit .env and .env.local:

APP_ENCRYPTION_SECRET=your-32-character-secret-here!!
DATABASE_URL=file:./leadflow.db

3. Run database migration

npx prisma migrate dev

4. Start the app

npm run dev

Open http://localhost:3000

---

API Keys — Where to Get Them

Go to Settings in the sidebar to enter and save credentials. All keys are AES-256 encrypted before storage.

Anymailfinder

• Sign up at anymailfinder.com
• Account → API → copy your key (format: starts with your account prefix)
• Pricing: 2 credits/valid decision maker email, 1 credit/person email
• Bulk API charges only when downloading results

Apify

• Sign up at apify.com
• Console → Account → Integrations → API tokens
• The app uses the apify/google-search-scraper actor (pay-per-event)

Vayne

• Sign up at vayne.io
• Dashboard → API Settings → generate token
• Connect LinkedIn in the Vayne dashboard — Vayne manages the LinkedIn session on their end
• No need to manually export cookies

---

Pipeline Workflows

Tab 1 — AirScale → Email

1. Export companies from AirScale as CSV
2. Upload → map domain column
3. Select decision maker categories
4. Start Enrichment → bulk API runs asynchronously
5. Export CSV/Excel

Tab 2 — LinkedIn Sales Navigator → Email

1. Build search in Sales Navigator, copy URL
2. Paste URL + set max results → Start Scrape
3. Vayne scrapes profiles (polls until done)
4. Select profiles → Enrich with Emails
5. Export results

Tab 3 — SERP → Email

1. Enter search term (e.g. "Solaranlage Installateur Deutschland")
2. Set country/language, enable social filter
3. Select decision maker categories
4. Start → Apify scrapes Google, domains extracted, then bulk enriched
5. Export results

---

How Anymailfinder Bulk Works

1. All domains submitted as one POST to /v5.1/bulk/json
2. Poll status every 5s until completed
3. Download results once (credits charged at download, not submission)
4. Speed: ~1,000 domains per 5 minutes

---

Database

SQLite at ./leadflow.db. Inspect with:

npx prisma studio

---

Deploy with Coolify (Docker)

1. Add the Gitea repo in Coolify

• New Resource → Application → Docker Compose (or Dockerfile)
• Repository: https://gitea.onyva.dev/TimoUttenweiler/lead-scraper
• Branch: main
• Build method: Dockerfile (recommended) or Docker Compose

2. Set environment variables in Coolify

In the app's Environment Variables tab:

Variable	Value
APP_ENCRYPTION_SECRET	Any random 32-character string (generate once, keep secret)
DATABASE_URL	file:/data/leadflow.db
NODE_ENV	production
PORT	3000

3. Add a persistent volume

• Mount path in container: /data
• This stores the SQLite database across deployments

4. Port mapping

• Container port: 3000
• Coolify will proxy via its reverse proxy automatically

5. Deploy

Click Deploy — Coolify builds the image from the Dockerfile, migrations run automatically on startup via docker-entrypoint.sh.

Dockerfile stages

1. deps — npm ci + native build tools for better-sqlite3
2. builder — Prisma generate + next build (standalone output)
3. runner — minimal Alpine image, non-root user, copies only what's needed

---

Troubleshooting

Issue	Solution
"API key not configured"	Add key in Settings
Job stuck at "running"	Check server console (npm run dev terminal)
Prisma errors on build	Run npx prisma generate && npm run build