ERP Integration

Eryxon Flow supports two-way data synchronization with external ERP systems like SAP, NetSuite, Odoo, and others.

Architecture Overview

flowchart LR
  ERP["Your ERP"] -->|REST sync, bulk-sync| API["Edge Functions<br/>(API key)"]
  API --> PG[("Postgres<br/>external_id · sync_hash · RLS")]
  PG -->|triggers| WH["Signed webhooks"]
  WH --> ERP

Integration Options

1. REST API Sync (Real-time)

Best for: Webhook-driven updates, event-based sync, real-time integration.

Method	Endpoint	Use Case
PUT	/api-{entity}/sync	Upsert single record
POST	/api-{entity}/bulk-sync	Batch upsert (up to 1000)
DELETE	/api-{entity}?id={uuid}	Hard delete

Sync endpoints exist for jobs, parts, operations, batches, cells, resources, materials, time entries, webhooks, and assignments. See the REST API reference for the full catalog.

2. CSV Batch Import (UI-based)

Best for: Initial data migration, periodic bulk updates, manual imports.

Navigate to Admin → Data Import in the web UI.

Supported Entities

ERP Concept	Eryxon Entity	Sync Endpoint
Sales Order	Job	/api-jobs/sync
Work Center	Cell	/api-cells/sync
Equipment / Tooling	Resource	/api-resources/sync

Parts and operations are synced via the unified /api-erp-sync endpoint or created nested within jobs.

Key Fields

external_id

Every synced record should include:

{
  "external_id": "SO-12345",      // Your ERP's unique identifier
  "external_source": "SAP",        // Source system name
  ...other fields
}

The combination of (tenant_id, external_source, external_id) forms a unique constraint for upsert operations.

sync_hash

Eryxon automatically generates a SHA-256 hash of the payload for change detection. On subsequent syncs, unchanged records can be skipped.

synced_at

Timestamp of the last successful sync, useful for incremental updates.

Related Documentation

• CSV Import - Step-by-step CSV import instructions
• REST API Reference - Full endpoint documentation

Database Schema

The sync infrastructure adds these columns to core tables:

-- Added to jobs, parts, cells, resources
external_id TEXT,              -- ERP identifier
external_source TEXT,          -- Source system name
synced_at TIMESTAMPTZ,         -- Last sync timestamp

-- Soft delete support (jobs, cells)
deleted_at TIMESTAMPTZ,        -- NULL = active
deleted_by UUID                -- User who deleted

Best Practices

1. Always include external_id - Required for upsert operations
2. Sync dependencies first - Sync cells before operations that reference them
3. Use bulk-sync for batches - More efficient than individual requests
4. Implement webhooks - Receive real-time updates back to your ERP
5. Handle errors gracefully - Bulk responses include per-record errors