INDUSTRY JOURNEY

Pharma Factory

Batch 112A just failed compliance. No one can explain why.

In a regulated plant, the question is not only "what did the system show?" but "how was that answer produced?". This journey builds monitoring where every answer comes from a saved query anyone on the team can read.

The challenge

Pharmaceutical manufacturing demands:

• Traceability - every batch, every reading, every decision
• Audit-ready data - inspectors need proof, not promises
• Rapid response - a compliance failure has to be explained quickly
• Cross-system visibility - data scattered across historians, SCADA and MES

What OpenX gives you

Capability	How it helps
Queries Azi proposes	Plain language to KQL, accepted by a person
A record of changes	Once Provenance is set up in your landing zone, Saves to connections, surfaces and warm queries are recorded; accepted proposals are not (what is recorded)
Any AI over MCP	Connect Claude, ChatGPT or Cursor to the same saved queries
Access rights	Each member holds the rights their work needs

By default, Azi proposes and a person accepts. See You stay in the loop, and Greenfield for the mode where Azi's own proposals are accepted for you.

What you will build

Component	Description
Connections	Temperature, pressure and batch data from your systems
Surfaces	Cleanroom monitoring, QA gating, batch tracking
Warm queries	Compliance checks, threshold monitoring, trend analysis
API endpoints	REST endpoints for LIMS, MES or your own dashboards
MCP	Any MCP client reading the saved queries

Two tracks

Track A: Hybrid - add OpenX without replacing anything

You already run DeltaV, HighByte, SCADA or historians. This track adds OpenX beside them:

• Bring existing data in through a live-device connection
• Create surfaces that organize it
• Build compliance queries with Azi
• Call those queries over REST, with no change to your infrastructure

Start the hybrid track

Track B: Full - author it all

Starting fresh, or want complete control:

• Design your connection architecture from scratch
• Create a complete surface hierarchy
• Build a warm query library
• Give each team the access it needs

Start the full track

Before you start

Requirement	Details
An OpenX workspace	See Get a Workspace
Azure (optional)	Live devices use Fathym's shared IoT hub until your workspace has its own landing zone and IoT fabric in your Azure subscription
Data sources	At least one: a Simulator node, a historian, or sensors

Why it matters for compliance

Without saved queries	With saved queries
"The AI said it was compliant"	The query that produced the answer, there to read
Black-box recommendations	Logic anyone on the team can read
Scattered evidence	One workspace the whole team reads from

What OpenX records, and how it maps to data-integrity requirements, is on the Compliance matrix.

Choose your track

Next steps

• Other industries -> Industry journeys
• Common query shapes -> Query Patterns