Federal agencies are seeking new approaches and technologies to unlock the power of data through artificial intelligence and machine learning (AI/ML). Focusing on the fundamentals can help them empower the real change they seek. And do so during what could otherwise be an intimidating process. So when your organization thinks AI, first think about the basics; strategic focus, teamwork for data, and secure, data-driven architecture. By doing so you can chart a more successful path to implementation and leadership in artificial intelligence.
AI: Strategic focus
Before any data scientist unpacks their “AI/ML killer-apps” toolkit, they all start with questions about the data. Where is it coming from? What is its quality? How much cleaning does it need? What restrictions does it have? How do I access it? Can it run in my environment? Does someone else need access? And who needs the results?
Those questions normally begin at the tactical task level. But through the filter of delivering results within the constraints of their existing IT environment. But if you’re a Federal Agency Chief or CIO (or you work with one) and you’ve never had to confront those tactically-based data challenges, it might be a bit daunting to frame the strategic data imperatives to get your agency moving forward.
Add to that the tendency of the data scientists to focus on tools, killer-apps and GPUs. But the IT team is worried about infrastructure growth and management, cloud-expansion and cybersecurity. But it’s harder still when agency legal and compliance officers show up to remind the “Federal Executive” (that’s you) about the requirements of laws, policies, privacy and confidentiality surrounding the data.
AI: Teamwork for data
Before it is a “technology issue,” Digital Transformation for any organization, company or Federal agency requires a mission-focused team dedicated to achieving mission outcomes from the point of view of the organization’s data. At Cisco, we’ve helped numerous organizations across all industries and branches of Government with Digital Transformation efforts. We’ve even made sure our own Digital Transformation was top notch.
As a result of these experiences, we understand that it’s critical to involve mission owners, IT teams, data scientists, and data governance team members in the process. This helps create an architecture that can support desired mission outcomes, daily operations, security and compliance management.
AI: Secure, data-driven architecture
An Intent Based Network Architecture encompasses everything required to deliver an organization’s desired mission outcomes in a secure, end-to-end, effective, efficient and compliant way. For an AI/ML architecture, a comprehensive approach must consider the requirements of and on the data in relationship to the infrastructure through a “data-pipeline”. That pipeline should include:
- data preprocessing
- model training
- model validation
- And model deployment.
The architecture must also implement capabilities that support data input-output intensive operations at the edge. It should also support compute intensive requirements at the core and/or in cloud. Plus, latency-sensitive necessities at the edge and on-prem.
You should consider these characteristics throughout the data-pipeline (while the data is being collected, cleaned, correlated, trained and modeled). Translating the answers to those fundamental questions for implementation in an intent based architecture ensures that data will be ingested, tagged, transformed, stored, cataloged, made available and accessed within appropriate compliance requirements and frameworks.
The real key to leadership
Data scientists generally avoid spending their time managing workstations/servers or cloud infrastructure. Instead, they want agile access to data, compute resources, workloads and devices that are based on a Zero Trust architecture; one that enforces data governance requirements.
At the same time, the IT Department is trying to provide their internal customers with an affordable, secure, seamless experience between edge, on-prem and in Multicloud environments — one also protected by a Zero Trust enforced security policy. All the while data gravity dictates that network compute and transport requirements are optimally allocated. Also, that they are easily managed across the agency’s edge, on-prem and cloud resources.
And modern AI/ML workflows, like Kubeflow, require network infrastructure management tools that make it easy to securely employ containers and modern workflows between on-prem resources and in Multicloud environments (like Cisco’s Container Platform). With all these needs in play, it’s easy to get sidetracked.
Only Cisco can provide a validated, end-to-end, data-pipeline purpose-built AI/ML architecture that ensures optimally allocated NVIDA GPU-powered compute and storage resources on a data platform with unified management.
Such a platform powers the full AI data lifecycle required by data scientists and simultaneously demystifies AI/ML stacks for the IT Department from edge, on-prem and in the cloud. Agencies that empower their workforce with an Intent Based Network Architecture and validated solutions like Cisco’s Data Intelligence Platform, will realize greater value from data and lead the Government in delivering AI/ML-driven mission outcomes—while ensuring appropriate security and compliance with data governance laws and policies, via Zero Trust.
Join the conversation
Cisco is proud to be a Platinum Sponsor of this year’s NVIDIA GPU Technology Conference, the premier event on artificial intelligence. Be sure to visit our booth (#104) and we’ll bring you up to speed on the latest Cisco solutions and our expanding AI ecosystem of partners. Plus, be sure to join us for two exciting Cisco led sessions.
When: November 5-6
Where: Washington DC
Government attends for FREE
Social Media: #GTC19