echos-ocr/.env.example

# Database (shared PostgreSQL on CT 102)
DATABASE_URL="postgresql://echos_ocr:YOUR_PASSWORD@192.168.68.102:5432/echos_ocr"

# Vercel AI Gateway (routes to OpenAI, Google, Anthropic, etc.)
AI_GATEWAY_API_KEY=""

# Ollama (only needed if using Ollama as the AI provider)
OLLAMA_BASE_URL="http://192.168.68.108:11434"

# MinIO S3 Storage (CT 105)
MINIO_ENDPOINT="192.168.68.105"
MINIO_PORT="9000"
MINIO_ACCESS_KEY="minioadmin"
MINIO_SECRET_KEY="YOUR_MINIO_SECRET"
MINIO_BUCKET="echos-ocr"

# Folder Watch (optional, mount a host path into the container)
WATCH_DIR=""
Build Echo OCR app: full-stack response card scanner - Next.js 15 + Tailwind v4 + shadcn/ui + TanStack Table - Prisma + PostgreSQL schema for ResponseCard, ProcessingJob, AppSettings - Ollama vision integration with structured extraction prompts - PDF-to-image pipeline with pdf2pic + sharp - MinIO S3 storage for uploaded files and extracted images - Chokidar-based folder monitoring for automatic processing - REST API (9 endpoints) ready for Monday.com integration - Dashboard with filterable data table, chip filters, bulk actions, CSV export - Card detail page with editable fields and side-by-side scanned images - Upload page with drag-and-drop and real-time processing queue - Settings page for Ollama, folder watch, and Monday.com config - Dockerfile (multi-stage) + docker-compose for local dev - Configured for Coolify deployment at echo.stillwell.cloud Made-with: Cursor 2026-03-10 08:17:45 -04:00			`# Database (shared PostgreSQL on CT 102)`
			`DATABASE_URL="postgresql://echos_ocr:YOUR_PASSWORD@192.168.68.102:5432/echos_ocr"`

Switch to Vercel AI Gateway for multi-provider OCR Replace individual provider SDKs (@ai-sdk/openai, @ai-sdk/google, @ai-sdk/anthropic) with the Vercel AI Gateway. Cloud models are now accessed via a single AI_GATEWAY_API_KEY with provider/model strings (e.g. openai/gpt-4o-mini, google/gemini-2.5-flash). Ollama remains available as a local fallback via @ai-sdk/openai-compatible. Made-with: Cursor 2026-03-11 12:05:38 -04:00			`# Vercel AI Gateway (routes to OpenAI, Google, Anthropic, etc.)`
			`AI_GATEWAY_API_KEY=""`
Integrate Vercel AI SDK for multi-provider OCR Replace the raw Ollama fetch-based OCR with the Vercel AI SDK, adding support for OpenAI, Google Gemini, Anthropic, and Ollama as selectable providers from the Settings page. Uses generateText with Output.object() and Zod schemas for type-safe structured data extraction. Made-with: Cursor 2026-03-11 10:24:15 -04:00
			`# Ollama (only needed if using Ollama as the AI provider)`
Build Echo OCR app: full-stack response card scanner - Next.js 15 + Tailwind v4 + shadcn/ui + TanStack Table - Prisma + PostgreSQL schema for ResponseCard, ProcessingJob, AppSettings - Ollama vision integration with structured extraction prompts - PDF-to-image pipeline with pdf2pic + sharp - MinIO S3 storage for uploaded files and extracted images - Chokidar-based folder monitoring for automatic processing - REST API (9 endpoints) ready for Monday.com integration - Dashboard with filterable data table, chip filters, bulk actions, CSV export - Card detail page with editable fields and side-by-side scanned images - Upload page with drag-and-drop and real-time processing queue - Settings page for Ollama, folder watch, and Monday.com config - Dockerfile (multi-stage) + docker-compose for local dev - Configured for Coolify deployment at echo.stillwell.cloud Made-with: Cursor 2026-03-10 08:17:45 -04:00			`OLLAMA_BASE_URL="http://192.168.68.108:11434"`

			`# MinIO S3 Storage (CT 105)`
			`MINIO_ENDPOINT="192.168.68.105"`
			`MINIO_PORT="9000"`
			`MINIO_ACCESS_KEY="minioadmin"`
			`MINIO_SECRET_KEY="YOUR_MINIO_SECRET"`
			`MINIO_BUCKET="echos-ocr"`

			`# Folder Watch (optional, mount a host path into the container)`
			`WATCH_DIR=""`