(US & Canada) Fredy Diaz, Deputy CDO at the U.S. Department of Agriculture, speaks with Adita Karkera, CDO, Government and Public Services at Deloitte, in a video interview about how data strategy is helping the USDA, building AI strategy, operationalizing AI, open data work at USDA, balancing governance and open data democratization, data literacy approaches, and advice for fellow professionals.

At the outset, Diaz delves into how the data strategy at USDA is proving helpful as the department builds its AI strategy. He says that the USDA is developing the AI strategy in accordance with the Office of Management and Budget, and the credibility built through the creation and execution of the data strategy is paying dividends.

Adding on, Diaz shares that the department is repeating the exercise with different and new stakeholders. He notes that the data strategy has not just helped in crafting the AI vision for USDA but has also enhanced the ability to execute newer technologies like generative AI.

Further, Diaz remarks that while strategy may not be a priority for everyone, it is an area worth investing in, as it can ensure scalability.

When asked how the USDA operationalizes AI, he shares two examples. First, Diaz mentions the AI Center of Excellence at the USDA. He adds that what started out of the scientific and research community has expanded across USDA and is now part of its governance structure.

Consequently, the department now has practitioners who are called upon to resolve AI issues. The next best thing, according to Diaz, is the USDA’s collaboration with the Chief Technology Officer and the ability to deliver on AI.

To quickly spin up technologies, the department has been working on incubators, not just for AI but for other projects as well. He acknowledges the fruitful partnership and expects that when it fully matures, it will reduce technology evaluation time, leading to a faster reaction to landscape changes.

Moving forward, Diaz discusses the open data work at USDA and managing data as an asset. He elaborates that the availability of an enormous amount of public data has led to the development of many public data products, including the flagship product, the Census of Agriculture.

Additionally, there are other products that USDA hosts on its open data platform, says Diaz. One such product is the “rural data gateway,” which allows stakeholders to see how world development investments are made across the country.

Apart from that, Diaz mentions publishing data in a classroom. Starting with a few universities, the goal is to ensure that students have a real-world data set and problem statement to work on.

While it generates valuable feedback from students and professors, it also acts as an implicit talent acquisition tool, says Diaz. After leveraging it, students may want to consider a career in public service, he adds.

According to Diaz, there is a paradigm shift, with not only students but also USDA employees wanting to make the data public and use it for such purposes. He remarks that data is shifting from a check-in-the-box requirement to a strategic asset.

Commenting on balancing governance and open data democratization, Diaz stresses having proper processes in place. Many times, it boils down to what is fit for use, and with sharing dashboards publicly, the USDA has seen a positive outcome.

To balance the governance aspect, Diaz mentions removing personally identifiable information and states that the USDA has a solid data governance process. He adds that the assistant CDOs are authorized to make governance-related decisions on behalf of the mission area. As CDOs, they must ensure that the concerns of the program are addressed while still freeing up the data asset.

Speaking of data literacy approaches, Diaz refers to three relevant programs initiated by the USDA:

The USDA data science program
Rotational programs at USDA
Rapid prototyping and hackathon events

Starting out of the food and nutrition service and led by an Assistant CDO, this program started with two dozen participants and now has a little less than 200. Through the rotational programs, employees get hands-on experience across mission areas.

The last approach will provide opportunities for employees to build something with what they have learned from the other programs.

Next, Diaz advises people to volunteer and take advantage of the rotational opportunities that the organization has to offer. He maintains that everyone should make their intentions known, show up, and say what they are good at.

In conclusion, Diaz urges emerging data professionals to embrace data challenges that are public in their organizations. He recommends taking up such opportunities, displaying skills, and building a portfolio of work.

CDO Magazine appreciates Fredy Diaz for sharing his insights with our global community.

PODCAST | Good Quality Data Is Critical to Get the Feet Wet With GenAI — U.S. Venture VP Data and Analytics