Scorecard OLD

Assess Your Data Stack for AI Readiness

AI readiness starts with a candid look at your data ecosystem. This quick assessment will help you identify gaps, highlight strengths, and uncover opportunities to optimize your data infrastructure for AI success.

Need Help?
If you'd like help filling out the questions or reviewing your results, our team can guide you. Please reach out to us: info@tmdc.io or schedule a meeting with our team.

Press Enter

Step 1/5

How to use this Scorecard

This scorecard presents a series of questions, each rated on a scale of 1 to 10. For each question, select the rating that best reflects your current capability:

1 (Ad Hoc): No or minimal capability in place.
5 (Reactive): Moderate capability, but with gaps in coverage or manual processes.
10 (Purpose-Driven): High capability with comprehensive, automated, or systematic implementation.

Once completed, your responses generate a total score out of 150, providing a clear snapshot of your current state and a starting point for planning improvements.

Press Enter

Step 1/5

Lineage1. How quickly and accurately can you trace end-to - end data lineage from source systems to AI outputs , and use that lineage to assess the impact of any change?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

‍1 (Ad Hoc): Raw data with no lineage visibility. Impact analysis is manual, updates take weeks with a high level of rework.‍
5 (Reactive): Partial lineage documented with basic tagging and categorization, but coverage is incomplete and often manual. Impact analysis can take days, and at least 30% of changes require manual rework.‍
10 (Purpose-Driven): Comprehensive lineage tracking across all systems, connecting entities and relationships through knowledge graphs. Impact analysis and changes propagate the same day, almost automatically, with less than 5% rework.

Press Enter

Step 1/5

Parallel, Governed Workflows2. How integrated and governed are your analytical workflows, and how rapidly and reliably can they run in parallel to deliver business-ready outcomes?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Workflows are fragmented across tools; governance is manual or absent; runs are sequential and slow; outcomes take weeks or months.
5 (Reactive): Some workflows are connected with partial governance; parallelism is limited; outcomes are delivered in days to weeks with frequent rework.
10 (Purpose-Driven): Workflows are fully integrated, end-to-end, with automated governance; parallel execution is standard, and outcomes are delivered in hours to days with minimal rework.

Press Enter

Step 1/5

Self-service Access3. How well is data access democratized while maintaining security/compliance, and how long does it typically take for a new user to gain access to data?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Data access is siloed, controlled manually, and often inconsistent. Days to weeks for granting access; duplicated extracts or shadow copies of data are common.
5 (Reactive): Broader access is available, but security/compliance controls are applied reactively and unevenly. Most approvals are received within 1-3 days; some may require duplicate extracts.
10 (Purpose-Driven): Self-service access available to all authorized users/ agents with fine grained controls ensuring security and compliance. Access grants in minutes or less; no redundant copies.

Press Enter

Step 1/5

IT dependence4. Can business and Al teams quickly access enterprise data independently, without relying on IT?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Access is through IT, requiring tickets and manual interventions. IT-managed handoffs; multi-day cycles per request.
5 (Reactive): Some self-service access exists, but teams often still depend on IT for delivery.
10 (Purpose-Driven): Teams can directly access and activate trusted data products through self-service tools. Zero-ticket onboarding; governed endpoints available same day.

Press Enter

Step 1/5

Context Aware5. Can your systems recognize and differentiate the contextual meaning of the same data across use cases, with definitions that are co-owned and aligned between business and technical teams?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): No recognition of context; data is interpreted differently in each use case, with no business and technical alignment.
5 (Reactive): Some contextual differences are documented, and a few shared definitions exist, but ownership is fragmented and inconsistent.
10 (Purpose-Driven): Context is consistently recognized across use cases, with business and technical teams co-owning definitions.

Press Enter

Step 1/5

Semantic Layer6. Do you have a semantic layer that connects business concepts to data with traceability for compliance, and how quickly can new definitions be implemented?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): No semantic layer; business concepts and technical data are disconnected, with no traceability. Changes roll out per tool, over weeks.
5 (Reactive): A partial semantic layer exists, with some business concepts linked to data and limited lineage documentation. New metrics reach priority tools in days.
10 (Purpose-Driven): A comprehensive semantic layer links all business concepts to technical data, with automated lineage ensuring full compliance and auditability. Publish once goes live across BI/APIs in hours.

Press Enter

Step 1/5

Data Preparation Effort7. How much of your data preparation effort is spent on repetitive, manual tasks compared to automated reuse?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Nearly all data preparation is manual and custom for each project. 70-90% manual per project; heavy one-off pipelines.
5 (Reactive): Some reusable data products reduce repetitive work, but most preparation is still case-by-case, more than 30% manual work
10 (Purpose-Driven): Preparation is largely automated using production- grade, reusable data products across projects, less than 10% manual.

Press Enter

Step 1/5

Versioning and Cataloging8. Is your data versioned, cataloged, and delivered as organized assets that Al systems can consume without extra preparation, and how fast can a new use case be spun up?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Data is raw and unorganized, requiring extensive manual cleaning or transformation before Al can use it. Each use case needs weeks of prep.
5 (Reactive): Some data is catalogued and reused for specific use cases, but versioning and consistency are limited. New use cases take days to weeks.
10 (Purpose-Driven): Data is versioned, catalogued, and delivered as Al- ready assets that are composable and immediately usable across domains. Fully versioned/cataloged, Al-ready assets, new use cases in hours to days.

Press Enter

Step 1/5

Global Metrics9. Do you maintain consistent global metrics across teams and use cases, and how quickly can you correct metric drift?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Each team defines and calculates metrics independently, leading to discrepancies. Drift persists for weeks.
5 (Reactive): Some shared metrics exist for common use cases, but gaps and inconsistencies remain. Drift detected/corrected in days.
10 (Purpose-Driven): A unified catalog of standardized, global metrics is consistently applied across all teams and systems. Drift is auto-detected, corrections propagate in hours.

Press Enter

Step 1/5

Business-Aligned Data Assets10. Are your data assets designed for specific business outcomes rather than generic consumption?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Data is stored generically, with no alignment to business outcomes.
5 (Reactive): Some data assets are tailored to use cases, but most remain generic.
10 (Purpose-Driven): Data assets are engineered as outcome-driven products, reusable and composable for multiple business needs.

Press Enter

Step 1/5

Al Explainability11. How effectively can you audit Al decision paths-such as model inputs, outputs, and interpretability-to ensure regulatory compliance?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): No visibility into Al decision paths, model inputs/outputs, or interpretability.
5 (Reactive): Some audit logs and partial traceability of model inputs/ outputs exist, but interpretability is limited and not standardized.
10 (Purpose-Driven): Full transparency with traceable model inputs/ outputs, clear interpretability of decisions, clear documentation, and audit-ready compliance reporting.

Press Enter

Step 1/5

Al Agent Data Access12. Can Al systems reliably access data through readily available APls, such as REST, SQL, GraphQL endpoints, and MCP, and how quickly can they be exposed?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): Al systems require custom integration for each data source, with no standard APls. Weeks to expose new endpoints.
5 (Reactive): Some APls are available, but coverage is limited and not optimized for Al. Days to expose.
10 (Purpose-Driven): A comprehensive Al-native execution layer exposes consistent, well-documented APls (REST, SQL, GraphQL, MCP) across all data domains. Endpoints can be exposed in hours.

Press Enter

Step 1/5

Dev Tools Integration13. Can both technical teams and business applications access production data securely and directly through notebooks (such as Jupyter), development tools, or APls for rapid delivery of Al use cases, and how long does environment setup take?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): No direct access; teams must rely on exports or IT-managed pipelines. The setup takes weeks.
5 (Reactive): Some notebook integrations exist, but coverage is inconsistent and limited to technical users. Set up in days.
10 (Purpose-Driven): Full support for direct, secure access to production data for both technical and business teams via notebooks, APIs, and modern development tools. Environment Setup in hours

Press Enter

Step 1/5

Real-time Access14. Does your infrastructure support real-time data access for Al inference with minimal latency?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): No support for real-time access; data must be batch- processed before use. Seconds-to-minutes latency is unacceptable for inference.
5 (Reactive): Some workloads support near real-time queries, but latency is inconsistent. 95th percentile latency seconds.
10 (Purpose-Driven): Infrastructure is optimized for low-latency, real-time Al inference across all workloads. 95 / 99th percentile latency seconds across critical workloads.

Press Enter

Step 1/5

Data Quality Monitoring15. Do you have real-time monitoring and alerting for data quality issues to proactively prevent cascading model failures in Al systems?

1

2

3

4

5

6

7

8

9

10

Please select your response on the 1–10 scale.

1 (Ad Hoc): No monitoring; issues found manually or only after failures.
5 (Reactive): Some monitoring exists, but alerts are inconsistent, delayed, or need manual fixes.
10 (Purpose-Driven): Automated, real-time monitoring with proactive alerts, anomaly detection, and self-healing to prevent cascading failures.

Press Enter

Step 1/5

Finally, please enter your email to see the final score.

Please use your corporate email.

Press Enter

Step 1/5

Step 1/5

🎉
Your Score is 150

0-80 points

Early stage, with significant gaps to address.

81-130 points

Progressing, with strong capabilities and room to strengthen further.

131-150 points

Strong foundation, purpose-driven, and Al-ready.

One of our experts will reach out to work with you on translating your score into a clear, actionable roadmap for strengthening your data foundation and advancing AI excellence.

Get the AI Readiness Whitepaper

Oops! Something went wrong while submitting the form.

© 2025 The Modern Data Company