Skip to main content

DPCH08 — SLA–SLO Backed & Observable

“Observable with SLAs and Quality Guarantees”

What DPCH08 is really asserting​

DPCH08 is not asserting that:

“Dashboards exist” or “metrics are collected.”

It is asserting that:

The Data Product is continuously observed, evaluated, and governed as a living product — including automated, enterprise-level Product Maturity Driven Development (PMDD) tests that monitor its productization health over time.

Observability here includes operational health and productization maturity health.

The Essence (HDIP + PMDD Interpretation)​

A Data Product satisfies DPCH08 if and only if:

  1. Its runtime behavior is observable (availability, freshness, latency, quality)
  2. Its productization maturity is continuously assessed
  3. Both are evaluated through automated, platform-level mechanisms

If maturity assessment is:

  • manual,
  • ad hoc,
  • spreadsheet-based,
  • or audit-driven,

then DPCH08 is not met, even if operational metrics exist.

Role of PMDD Enterprise Tests​

PMDD enterprise tests are the governance observability layer of HDIP.

They are:

  • always-on
  • automated
  • periodically executed (e.g. weekly)
  • independent of domains

In DPCH08 terms:

PMDD tests observe whether the Data Product continues to meet the characteristics of a product — not just whether it runs.

This elevates maturity from a one-time assessment to a continuously monitored property.

Positive Criteria — When DPCH08 is met​

DPCH08 is met when all of the following are true:

1. Operational SLAs / SLOs are defined and observed​

The Data Product declares and measures:

  • availability
  • freshness / timeliness
  • latency (where applicable)
  • quality thresholds

Metrics are:

  • time-stamped
  • observable by consumers and governance
  • tied to the product, not just pipelines

2. PMDD maturity tests are in place and automated​

The enterprise runs PMDD tests that:

  • evaluate DPCH characteristics (01–14)
  • consume signals from HDIP shared services
  • produce periodic maturity snapshots
  • detect regressions in productization

This is a platform responsibility, not a domain task.

3. Maturity observability is continuous, not episodic​

Evidence includes:

  • scheduled maturity evaluations (e.g. weekly)
  • historical maturity trends
  • last-evaluated timestamps per DPCH

A maturity score frozen in time does not satisfy DPCH08.

Negative Criteria — When DPCH08 is not met​

DPCH08 is not met if any of the following are true:

❌ Observability is runtime-only​

Examples:

  • dashboards show pipeline health
  • SLAs exist, but no productization visibility
  • maturity assessed once during onboarding

This observes systems, not products.

❌ PMDD tests are manual or optional​

Examples:

  • maturity checklists filled once per quarter
  • governance reviews without automation
  • self-attestation without verification

Manual governance does not scale and fails DPCH08.

❌ Observability belongs to engineering, not governance​

Examples:

  • metrics visible only to platform teams
  • maturity hidden in internal reports
  • no governance-facing dashboards or alerts

DPCH08 requires shared visibility.

Edge Cases (Important Guidance for Agents)​

Case 1: “SLAs exist, but no maturity monitoring”​

⚠️ Partial

Rationale:

  • operational maturity exists
  • productization maturity is unmanaged
  • common transitional state

Case 2: “Maturity monitored manually, quarterly”​

❌ Not met

Rationale:

  • episodic assessment ≠ observability
  • violates PMDD principles

Case 3: “Weekly PMDD maturity snapshots + SLO dashboards”​

✅ Met

Rationale:

  • both runtime and productization health are observable
  • maturity becomes a governed signal

Evidence Signals an Agent Should Look For​

Authoritative evidence:

  • PMDD maturity snapshot dataset
  • scheduled maturity evaluation jobs
  • maturity trend dashboards
  • timestamps for last DPCH evaluation

Supporting evidence:

  • SLA/SLO dashboards linked to product
  • alerting rules for regressions
  • governance portfolio views

Red flags:

  • maturity scores stored in documents
  • manual governance sign-offs
  • no history or trend data

How an AI Agent Should Decide​

Decision rule (simplified):

If the enterprise cannot continuously observe both the operational health and the productization maturity of the Data Product through automated PMDD tests, DPCH08 is not met.

Why DPCH08 Is Non-Negotiable​

Without DPCH08:

  • maturity degrades silently
  • regressions go unnoticed
  • governance becomes reactive
  • Data Mesh loses systemic integrity

DPCH08 is where PMDD becomes real, turning maturity from intent into an observable system property.

Canonical Statement​

DPCH08 is satisfied only when a Data Product is continuously observable through defined SLAs/SLOs and automated Product Maturity Driven Development (PMDD) enterprise tests that monitor and trend its productization maturity over time.