Data Analytics & Engineering for SYSPRO
Phalanx is a metadata-driven ETL framework that turns a SYSPRO ERP database into an analytics-ready Postgres data warehouse — typically over a weekend, not a quarter. It reads SYSPRO's own Dds* data dictionary (DdsTables, DdsColumns, DdsForeignKeys, DdsColumnValues) to auto-discover the schema, so there are no hand-mapped column lists and no manual type coercion. From there it generates typed dbt staging models and keeps the warehouse current overnight with watermark-based incremental sync that only moves delta rows.
Ten core SYSPRO modules are covered out of the box — Inventory, Accounts Receivable, Sales Orders, Purchase Orders, General Ledger, Accounts Payable, Sales Analysis, Work in Progress, Bills of Material, and GRN — landing in dimensional marts (dim_/fct_) ready for BI tools.
For the analytics layer, Pharos exposes the warehouse (and live SYSPRO data) to AI assistants via the Model Context Protocol, so teams query KPIs, inventory health, and customer risk in plain English instead of SQL. Phathom adds process mining on top: reconstructing how orders, approvals, and jobs actually flow through SYSPRO, and where they deviate.
What it covers
Metadata-driven schema discovery
Phalanx reads SYSPRO's Dds* dictionary on first run to discover tables, columns, types, and foreign keys automatically. No hand-mapped schemas.
Watermarked incremental sync
Only delta rows move per cycle, so the warehouse refreshes overnight without hammering the SYSPRO server.
Auto-generated dbt models
Typed dbt staging models per module (e.g. `phalanx generate --module INV`), so analysts build dimensional marts without leaving dbt.
Ten SYSPRO modules out of the box
Inv, Ar, Sor, Por, Gl, Ap, Sal, Wip, Bom, and Grn — covering inventory, sales, purchasing, finance, and production.
AI-powered analytics via Pharos
Natural-language KPI dashboards, inventory health, and customer risk over the warehouse through the Model Context Protocol.
Process mining via Phathom
Reconstruct real execution paths from SYSPRO event logs: bottlenecks, deviations, and drift from designed process.
Frequently asked questions
- How do I get SYSPRO data into a data warehouse?
- Phalanx extracts SYSPRO into Postgres using SYSPRO's own Dds* metadata dictionary to auto-discover the schema — tables, columns, types, and foreign keys — with no hand-mapping. Initial load runs in bulk; after that, watermark-based incremental sync moves only changed rows, typically on an overnight schedule.
- Does Phalanx require a data engineering team?
- No. Phalanx is designed for manufacturers who have outgrown spreadsheet exports but can't justify full data-team headcount. Schema discovery, extraction, and dbt staging models are generated automatically; an analyst comfortable with SQL and dbt can build marts from there.
- Which SYSPRO versions and modules are supported?
- Phalanx is designed for SYSPRO 8 installations and covers ten core modules out of the box: Inventory, Accounts Receivable, Sales Orders, Purchase Orders, General Ledger, Accounts Payable, Sales Analysis, Work in Progress, Bills of Material, and GRN.
- Can I query SYSPRO data with AI or natural language?
- Yes. Pharos is an MCP (Model Context Protocol) server that gives AI assistants like Claude structured access to SYSPRO data and the Phalanx warehouse. Teams ask questions in plain English — inventory health, customer risk, MRP shortages — and get answers grounded in live data.
- Does extraction impact the production SYSPRO server?
- Impact is minimised by design: after the initial load, watermarked incremental sync touches only delta rows per cycle and typically runs overnight, outside production hours.
Running SYSPRO? Let's talk.
We integrate with your existing SYSPRO installation — no rip and replace. Book a discovery call and we'll map the fastest path to value for your operation.