Thinklytics

Texas Tech University System · Higher Education · Austin, TX · 16 weeks

Texas Tech University System AI Enablement case study

A state university’s research division needed a certified data infrastructure to secure $6.8M in federal AI grants. We audited data systems across 14 departments, found critical gaps, and fixed them within 16 weeks. This ensured the university met grant requirements and unlocked the funding.

Challenge

The university received $6.8M from the National Science Foundation to fund AI research, but only if it certified its research data infrastructure within 16 weeks. The certification covered 14 research departments. An internal review found 31 gaps in the existing infrastructure that needed fixing to meet the deadline.

Approach

We identified 31 certification gaps and prioritized them based on impact. Then, we created a focused remediation plan targeting the most critical issues first. To support this, we implemented a research data management platform that centralized data cataloging, enforced access controls, tracked data lineage, and documented reproducibility across all 14 departments.

Outcome

We closed all 31 gaps by week 14, two weeks before the certification deadline. NSF granted the certification, unlocking $6.8M in funding. Now, 340 researchers across 14 departments actively use the research data management platform.

How We Dug Into Sorting Issues by Certification Status and Got Results

So, we spotted 31 gaps overall. Not all of them were urgent, though. We got together with the NSF program officer to figure out which 8 needed fixing right away and which 23 could wait. Then, we tackled those 8 critical issues and wrapped them up in just six weeks.

How We Broke Down Our Process to Hit Reliable Results Every Single Time

The certification meant we had to map out every single research data pipeline, something nobody on the team had done before. So, we created a simple, no-fluff framework for documentation. Then, we trained 14 data coordinators to keep it current. That simple system didn’t just help us get certified; the NSF reviewer actually called it a model worth copying.

How we created a platform that keeps going strong long after the grant cash runs out

We put together the research data platform with open-source tools so the university wouldn’t be locked into one vendor. Then, we walked their IT team through how to handle maintenance and updates themselves. That way, the platform kept humming along long after the grant ran out. It actually ended up becoming the blueprint for all their future research data systems.

Results

  • $6.8M Federal grant funding secured
  • 31 Infrastructure gaps resolved
  • 14 Research departments certified
  • 340 Researchers on the platform

We had $6.8 million in grants tied to a certification deadline, and we weren’t sure we’d meet it. Thinklytics helped us close all 31 gaps with two weeks left. The funding went through, and their platform is much better than what we used before.

Vice Provost for Research, State University Research Division

In the client's terms

The question

Is the data in a state that will survive what we want to build on it?

The situation

The NSF had awarded $6.8M for AI research conditional on certifying research data infrastructure across 14 departments within 16 weeks. An internal review had found 31 gaps.

What we found

Thirty-one gaps and sixteen weeks meant the order of work decided whether the money arrived. Several gaps were shared across departments, so fixing them once cleared multiple findings, and that was not visible from the list.

What it produced

All 31 gaps closed by week 14, two weeks inside the deadline. NSF certification was granted and $6.8M in funding was released. Three hundred and forty researchers across 14 departments now work on the platform.

The roadmap

The 31 gaps were ranked by what they blocked rather than by severity, so the shared ones went first. A research data management platform was put in behind the remediation to handle cataloguing, access control, lineage and reproducibility documentation across all 14 departments, which is what stops the gaps reopening after certification.

The work behind this result

Related case studies

Frequently asked questions

What does NSF research data certification require?

A documented map of every research data pipeline, plus cataloguing, access control, lineage and reproducibility documentation. Texas Tech University System had 31 gaps against that across 14 departments and closed them in 14 weeks, releasing $6.8M in funding.

How do you prioritise infrastructure gaps against a funding deadline?

By what they block, not by severity. Working with the NSF programme officer here established which 8 of the 31 gaps had to close for certification and which 23 could follow. The critical 8 closed in six weeks.

How do you keep a research data platform running after the grant?

Build it so the institution can own it. Texas Tech's platform was assembled from open-source components and their own IT team was trained to maintain it, and it has since become the template for their other research data systems.