The Department of Veterans Affairs is continuing to see returns on its DevSecOps and product line management transformation.
One of its efforts has been with its Data and Analytics Product Line, which collects various data investments across the Office of Information and Technology (OIT) and aligns them in a single management line to share data resources across the enterprise.
Director of this product line, Joe D'Auria, helped develop an "execution plan" to merge these capabilities into a small number of platforms.
"Product engineering is spearheading many of the [software-as-a-service] adoptions, low code, no code build outs. Then, in OIT, exporting those experiences and providing those platforms to the rest of the enterprise,” D’Auria told GovCIO Media & Research.
This effort included migrating workloads to the cloud, like the product line team's data and analytics platform Rockies and the Summit data platform, both of which will provide shared services of cloud platforms securely.
“We’re doing what we can as product managers to create consumable data products once and let the enterprise consume them from those platforms,” D’Auria said. “We want all these data platforms to work seamlessly together using the same permission sets, so we're federating our security model so that we understand who has access to data.”
The data product line has also developed workflow-based frameworks and leveraged automation to improve data flow, access and tool adoption. VA's data catalog initiative is taking an enterprise-wide view of data assets to better understand authoritative data and explain how to access information.
“We want to make [data] consumable. And we want to make it digestible for everybody from developers who need to use our APIs to end users such as data analysts, data scientists that are going to build insights and improve veteran services,” D’Auria said.
VA’s Rockies and Summit platforms show the trusted sources for data and enable people to request access to it through mechanisms on those platforms. Having a consolidated data platform speeds adoption as opposed to manually discovering a raw data set or semi-curated dataset.
The product line is also collaborating with VA Chief Data Officer Kshemendra Paul and the Data Governance Counsel to drive the notion of data stewardship at the agency. Together, they work to identify key authoritative data sources, identify business owners, then load the system into the analytics environment. This process has enabled VA to improve enterprise-wide reporting by providing lineage from transaction to analytic insights.
“This really kind of changes that paradigm to where we know where the good data is. We've got structures and frameworks around it to bring it into a workspace,” D’Auria said.
The team is looking to adapt platform-as-a-service (PaaS) and software-as-a-service (SaaS) and use managed platforms to integrate those capabilities. VA is upskilling its workforce to better support these capabilities.
"We're automating those processes and not putting a lot of hands on the keyboard for it to data visualizations with low code platforms,” D’Auria said. “We look at those types of tools. We also look at the odds from a resiliency and survivability perspective.”
To drive resiliency, VA is leveraging proactive monitoring to ensure survivability of data assets and that the services on top of that data are highly available for end user consumers and is also honing in on architectural decisions to ensure that data pipelines are survivable and can adapt to change.
“You're reducing risks because you know where your data is, you know what the protections are, and you're inherently getting it in multiple data centers without even really thinking about it much more than saying that you need to,” D’Auria said.
Looking ahead, one of the team's biggest priorities will be to continue to standardize its cloud platform around Summit to pull data from any location, curate it, then make it available to VA and its strategic partners. The team will also be moving some of its data verticals and work streams to Summit to provide a shared services platform component, which will drive more efficient use of both federal "full-time-equivalent" employees and contractor personnel.
“We want to use Summit to make progress into our platform interoperability efforts. ... We don't want any of our data locked in silos, and that's why we're driving that interoperability piece,” D’Auria said. “We're trying to [modernize] in such a way that five years from now, we have less systems under management and [the data product line] — core common platforms that can talk to each other — and we can be focused on helping people create data insights, rather than building new platforms continuously.”