Curation & Observability
Improve metadata coverage and systematically close gaps across domains.
A multi-year transformation, sequenced for momentum
12–24+ months from baseline to scale. Each phase delivers usable governance — the DG Scorecard MVP is ready and curation efforts are already underway.
Assess Readiness & Alignment
Key steps
- Evaluate technical and cultural maturity
- Identify pain points and existing capabilities
- Align vision, set expectations, secure executive sponsorship
- Address concerns proactively
DG Scorecard MVP shipped
Define Domains, Roles & Scope
Key steps
- Identify domain boundaries based on analytical needs
- Define critical roles: DPO, Domain Owner, Steward, Custodian, Trustee
- Select 2–4 MVP data products as lighthouse projects
Curation efforts WIP · RACI defined
Enabling Infrastructure
Key steps
- Implement core platform capabilities — catalog, virtualization, transformation, governance automation, observability
- Make build vs. buy decisions with vendor tools and custom integration
- Prioritize standardization and user experience
EDPM Track 1 · Snowflake tags · Lineage · Row Access Policies
Metadata, Lineage & SLAs · Launch DG Council
Key steps
- Define metadata standards: business, technical, operational, governance
- Implement automated lineage tracking
- Define SLAs for freshness, availability, completeness, accuracy, latency, support
- Initiate DG Council to gain momentum and alignment on AI Ready charter
Council launch automated lineage
Launch MVP (EDPM) & Expand Gradually
Key steps
- Launch MVP (EDPM) data products, gather feedback, measure success
- Scale incrementally with early adopters
- Continue education, automation, governance maturity
- Hold retrospectives and evolve platform & culture continuously
Multi-year scale across all functions
Current state · EDPM (Finance)
Where we stand today
5 working
7 gaps
Live today
Working well
- EDPM migration — TD → Snowflake
- EDPM metadata — Critical metadata captured as Snowflake Tags
- Data Discovery — Alation catalog
- Compliance — Securiti.AI scanning PII
- Compliance — GWC in-take assessments
Open gaps
Not working yet
- Trust · DMF Quality — Data products leverage independently; DG not involved; Quality framework / backend to drive accuracy
- Trust · Lineage — Temp tables require mitigation
- Risk · DE resource constraints — Competing priorities
- Trust · Access Controls — Row access policies pending
- Data Discovery · Marketplace — Data Products marketplace TBD · POC required
- AI Readiness · Provenance, Bias — Definition of AI Ready unresolved
- AI Readiness · Semantic Layer Knowledge Context — Track 2 scope
Same plan, three-phase abstraction
The same execution rolled up to three milestones for executive readouts.
01
Foundation
· Months 1–3
02
Operationalization
· Months 4–9
03
AI-Readiness Standards
· 10 months
What we’re driving right now
Concrete initiatives moving the strategy from slideware to operating reality — measured against clear targets.
Data Governance Scorecard
Drive every data product toward AI-ready enterprise data. MVP is live.
Data Governance AI Agent
Answers questions on data usage, policies, ownership, and more — on demand.
TD → Snowflake Migration
Migrate Teradata tables to Snowflake as the AI-platform foundation.
High-level KPIs
How we measure AI readiness
0–0
Lighthouse products
MVP data products as lighthouse projects
Bi-weekly
Council cadence
Curation status & scorecard review
0%
Explainability
AI models with documented context / lineage
↑
Adoption
Catalog & glossary search frequency
