AUTOMATED DATA PRODUCTS · AGENTIC AI · SUPPLY CHAIN
Goldync builds automated, governed data products for supply chain teams — from raw ERP exports to trusted gold tables.
Get StartedTHE PROBLEM
SAP and Oracle exports are full of duplicates, inconsistent supplier names, and missing fields. Every analyst cleans the same data differently.
A schema change upstream kills dashboards downstream. Nobody knows until a VP asks why last month’s spend looks wrong in the board meeting.
Procurement, Finance, and Operations all have their own version of supplier spend. Reconciling takes weeks. Decisions are made on stale numbers.
THE SOLUTION
Goldync delivers production-grade, governed data products on Databricks — with agentic AI handling schema mapping, quality monitoring, transformation, and documentation automatically.
HOW IT WORKS
Databricks Auto Loader
Continuous pull from ERP, Coupa, Ariba, supplier portals. Schema evolution handled.
Delta Live Tables — Silver
Dedup, standardise, validate. Bad records quarantined. DQ rules auto-generated.
AI Agents — Transform
Business rules applied. UNSPSC hierarchy, currency normalisation, supplier dedup.
Delta Sharing / SQL Warehouse
Gold tables with SLA. Accessible via BI tools, REST API, or Delta Sharing.
AGENTIC AI LAYER
Reads source schemas and auto-proposes Bronze→Silver column mappings using LLM reasoning. No manual mapping spreadsheets.
Detects volume drops, null surges, distribution shifts after every run. Self-heals known issues. Raises alerts for new ones.
Takes plain-language business rules and generates tested PySpark. Validates in sandbox before merging to production pipelines.
Auto-generates column-level lineage, glossary entries, and catalog documentation after every pipeline deployment.
FIRST DATA PRODUCT
gold_procurement.po_spend_analytics
| Column | Type | Description |
|---|---|---|
| po_key | STRING | Surrogate key (SHA-256 hash) |
| supplier_id / name | STRING | Normalised, deduped supplier |
| category_l1/l2/l3 | STRING | UNSPSC hierarchy |
| po_amount_usd | DECIMAL(18,4) | PO value in USD |
| 3way_match_status | STRING | MATCHED / PARTIAL / UNMATCHED |
| on_time_delivery_flag | BOOLEAN | GR vs. promised date |
| supplier_risk_score | FLOAT | AI-computed risk score 0–1 |
DOMAIN ROADMAP
PO spend · Supplier master · 3-way match
Stock levels · Reorder points · Aging
Shipment tracking · Carrier performance · OTIF
Forecast accuracy · Sell-through · Seasonality
Financial health · ESG scores · Concentration
THE IMPACT
Guaranteed freshness backed by monitoring
Gold tables updated every 4 hours or less
vs. traditional manual data engineering
Running quality, transform, lineage & docs
Schema Discovery Agent handles it all
From kickoff to production gold table
BUILT ON
Every Goldync data product runs natively on Databricks, leveraging the full power of the Lakehouse platform — on AWS, Azure, or GCP.
Databricks Auto Loader · CDC Connectors
Delta Live Tables (DLT) · PySpark
Delta Lake · Unity Catalog
Databricks Workflows · Event-Driven Triggers
SQL Warehouse · Delta Sharing · REST API
Claude API · LangChain · Great Expectations
Ready to build your first data product?
support@goldync.com