nuMetrix: RMC Material Analytics

Demo Walkthrough — March 2026

48,382 materials · 68,441 cases · 2.27M billing events
Real data from a Swiss hospital

What you're about to see

The system analyzes RMC's real data through multiple analytical lenses.

38
Analytical Signals
106K
High-Severity Findings
10
Root Causes Identified
10 / 11
Theses Confirmed

Every thesis with a confirmed verdict also has a verdict — a structured root cause with confidence score and recommendation. The system doesn't just count anomalies; it explains them.

Live demo step

Open JinDesk at localhost:4000. With only zeta built, it auto-redirects to hospital_zeta. Show the overview page with its summary cards.

Presenter: The overview page shows all this at a glance. Let the numbers sink in before moving on.

Let's zoom into one specific story.

The Catalog Maintenance Gap

Thesis: is_active flags and validity dates are drifting apart

Confirmed
Verdict
6,880
Findings
CHF 6.7M
Money at Risk
100%
Evidence Score

Four signals feed into this thesis — all four fired:

SignalRoleFindings
Active but date expiredPrimary1,368
Inactive but date validPrimary1,413
Stale articlesSupporting964
Catalog health perspectiveContext3,135

Live demo step

Navigate to Theses → click "Catalog Maintenance Gap". Show the evidence chain and confirmed verdict.

the viewer's Question

The question that changed the signal.

"OK, so we have 1,368 materials marked active with expired dates. But which field is actually wrong? The flag or the date? If the article is still being purchased, the date is probably wrong. If nobody's buying it, the flag is wrong."
— the viewer, nuMetrix Partner, March 2026

This is domain knowledge that no data model can derive. The system knew something was wrong, but the viewer knew what to do about it.

Presenter: Pause here. This is the key moment — the transition from "data anomaly" to "actionable insight." the viewer's classification logic is what makes the finding useful.

Activity-Based Classification

the viewer's rule: purchasing activity tells you which field is wrong.

Signal Has Purchases? Likely Wrong Field Action
Active + date expired Yes Date Extend validity — article genuinely in use
Active + date expired No Flag Set inactive — article genuinely expired
Inactive + date valid Yes Flag Re-activate — article genuinely in use
Inactive + date valid No Date Shorten validity — article genuinely inactive

Simple rule, massive impact. This logic is now built into the signals — every finding carries the classification automatically.

Presenter: Walk through the table slowly. The audience should understand the 2x2 matrix before seeing the numbers.

What the data says

Real RMC numbers after applying the viewer's classification.

Active + Date Expired (1,368)

ActivityWrong FieldCountCHF
Purchases Date 610 18,984
None Flag 758 92,706

65 findings escalated to high severity (active + purchases + CHF>1K = urgent date fix)

Inactive + Date Valid (1,413)

ActivityWrong FieldCountCHF
Purchases Flag 216 14,860
None Date 1,197 109,345

16 findings escalated to high severity (inactive + purchases = being bought under the radar)

Live demo step

Navigate to Findings → select signal_active_date_expired. Expand a finding to show the has_purchases, likely_wrong_field, and purchase_count evidence fields.

How did we get from conversation to code?

Knowledge Capture: Subject Matter

the viewer's insight — formalized, attributed, versioned.

business_rule active Level 0 — Observation

Activity-based classification of is_active/date contradictions

When an article's is_active flag contradicts its validity date range, purchasing activity in the observed period reveals which field is likely wrong. Active articles with recent purchases have a wrong date; active articles without purchases have a wrong flag.

Provider: The viewer Role: nuMetrix Partner Date: 2026-03-03
Anchored to: signal_active_date_expired signal_inactive_date_valid thesis_catalog_maintenance_gap

Live demo step

Navigate to Subject Matter → find "Activity-based classification of is_active/date contradictions". Show the provider attribution, anchors, and the "why" field.

Presenter: Emphasize: this is the viewer's knowledge, with their name on it. The system attributes everything. No anonymous rules.

The Knowledge-Driven Workflow

How expert knowledge becomes system intelligence — in one session.

1

Capture

The viewer shares their classification rule. It becomes a Subject Matter entry — attributed, scoped, versioned.

2

Consequence Analysis

Which signals need to change? What evidence fields to add? How does severity shift?

3

Implement

Signals are enriched with billing/usage joins. Activity classification built into every finding.

4

Rebuild & Review

Full rebuild across all 5 tenants. Same findings, richer evidence. 81 findings escalated to high severity.

From conversation to production: under 30 minutes. The knowledge is permanent — it survives every rebuild, every new tenant.

The Bigger Story: Dossiers

Subject Matter accumulates. Dossiers weave it into narratives.

Material Data Quality at RMC

A curated narrative connecting 5 Subject Matter entries about material master data quality at RMC — from wrong descriptions to validity contradictions to catalog maintenance gaps.

Curator: The viewer · 5 entries · active

Each Subject Matter entry is atomic — one insight, one provider, one scope. Dossiers are the story layer: "Here's what we know about material data quality at this hospital, and here's why it matters."

Live demo step

Navigate to Dossiers → open "Material Data Quality at RMC". Show the narrative and the linked entries.

the viewer has four more questions.

Catalog Inventory KPIs

the viewer asks four questions. Each one is now a signal — finding count IS the answer.

the viewer's QuestionSignalAnswer (RMC)
How many unique active articles? signal_active_materials 22,402
How many unique inactive articles? signal_inactive_materials 25,980
Active articles from a single vendor only? signal_active_single_source 487
Articles delivered by multiple vendors? signal_multi_source_materials 717

All four feed into perspective_catalog_inventory (48,382 materials assessed). The same pattern: question → signal → instant answer across any tenant.

Live demo step

Navigate to Findings → select signal_active_materials. The finding count in the header is the viewer's answer: 22,402 active articles. Switch to signal_active_single_source to show the 487 single-vendor materials with supplier names and billing volumes.

Presenter: The point: the viewer's questions are not one-off queries. They're signals that run on every rebuild, across every tenant. hospital_alpha shows 202 active, 78 single-source — same question, different hospital, instant answer.

The Report

the viewer's questions — answered, explained, printable.

nuMetrix — Catalog Inventory Report
A Swiss hospital
04.03.2026
48,382
Total
22,402
Active
25,980
Inactive
BucketCount
Active, single supplier487
Active, multiple suppliers717
Active, no supplier data (derived)21,198
Total active22,402
Lineage: materials.csv → bronze → silver → gold_materials + gold_material_sourcing → signals

Every number links to its signal. Every signal shows the SQL logic. Print to PDF or share directly.

Live demo step

Navigate to Executive Summary (sidebar) → click Catalog Inventory. Show the business questions table, supplier breakdown, and lineage section. Hit Ctrl+P to show the print-ready PDF.

Presenter: This is the "I can hand this to my boss" moment. The report is generated from live data, not a screenshot. Same report, different tenant, different numbers.

Search for "the viewer"

The JumpBar finds people, signals, theses, materials — everything in one search.

🔍 ronnie esc
People
⊙
the viewer
nuMetrix Partner · 8 entries · 2 Dossiers

Live demo step

Press Cmd+K (or J) to open the JumpBar. Type "ronnie". Show the People result with contribution counts.

Presenter: This is a nice "aha" moment. The viewer sees their own name in the system, attributed to their contributions. Also try searching "material", "validity", or "catalog" to show how everything is connected.

The big picture.

Confirmed Theses at RMC

10 out of 11 theses confirmed by evidence.

ThesisVerdictFlagged Volume (CHF)
Cost center misallocationdata quality66.7M
Cross-site leakageprocess failure66.5M
Stale catalogueprocess failure39.2M
I/O imbalanceprocess failure35.4M
Revenue leakage (unbilled)process failure34.8M
Duplicate billingdata quality33.3M
Supplier concentration riskstructural28.7M
Overpriced genericsstructural28.4M
Catalog maintenance gapprocess failure6.7M
Phantom casesdata quality33.3M

Each confirmed thesis has a verdict with root cause category, confidence score, explanation, and recommendation. The system doesn't just flag — it explains.

Live demo step

Browse the Theses page. Click into one (e.g., "Stale Catalogue") to show the verdict section with root cause and recommendation.

The jinflow

From raw data to actionable root causes — every layer adds meaning.

Verdicts
10 root causes
Theses
11 business questions · 10 confirmed
Signals & Perspectives
38 analytics · 531K findings
Gold Layer
48K materials · 68K cases · 2.3M billing events
+ 55 entries & 15 Dossiers: expert knowledge that enriches every layer

What happens next?

Every conversation adds knowledge. The system grows smarter with each Subject Matter entry.

What the viewer can do now

  • Share domain rules → they become Subject Matter
  • Challenge signal results → signals get sharper
  • Curate Dossiers → build the institutional narrative
  • Validate findings → calibrate confidence

What the system does

  • Attributes every insight to its provider
  • Versions everything — knowledge evolves
  • Runs across all tenants — portable rules
  • Rebuilds in minutes — instant feedback
"The difference between data and knowledge is the why."
— nuMetrix design principle
Presenter: End with this quote. The whole point is that the viewer's "why" is now in the system, attributed, and working across every rebuild.

Thank you

nuMetrix · Hospital Material Flow Analytics

55 entries · 15 Dossiers · 38 Signals · 11 Theses · 10 Verdicts
All knowledge attributed. All results reproducible.