Most enterprise AI projects fail after the demo because real-world data infrastructure is far messier, more fragmented, and more insecure than most companies realize.In this episode of An Hour of Innovation podcast, Vit Lyoshin speaks with David Bauer, co-founder and CTO of Axonis. David is an AI systems architect and expert in federated AI, enterprise security, and large-scale predictive systems.Vit and David explore why so many enterprise AI projects never reach production, how centralized data lakes are becoming “data swamps,” and why the future of AI may depend on securely bringing models directly to production data instead of moving sensitive data into centralized systems. The conversation also dives into AI agents, cybersecurity risks, predictive analytics, and lessons learned from building high-stakes AI systems for government and national security environments.They discussed:* Why enterprise AI fails after successful demos* The hidden “last mile” problem in enterprise AI* Why data lakes become unusable “data swamps”* Federated AI architecture for secure enterprise systems* AI agents and emerging cybersecurity risks* Why production data breaks most AI implementations* How sensitive enterprise data should stay decentralized* Predictive COVID analytics used by government agencies* AI infrastructure lessons from DARPA and national security* Why future AI systems won’t rely on promptsDavid Bauer is an AI systems architect focused on federated AI, enterprise security, and large-scale distributed intelligence systems. His work spans enterprise AI infrastructure, predictive analytics, secure data architectures, and agentic AI workflows operating in highly sensitive environments. Over the years, David has contributed to projects connected to national security, government AI initiatives, and advanced predictive modeling systems used for logistics and operational planning.Timestamps00:00 Introduction01:14 AI Challenges in Enterprises03:15 Understanding Last Mile AI Problem05:22 Exploring Federated AI and Its Benefits09:10 Security Considerations in Federated AI13:08 Implementing Federated AI: Trade-offs and Frameworks15:55 Real-World Applications of Federated AI19:12 Example of Using Federated AI21:16 How to Start with Federated AI25:29 Lessons from COVID Data Tracking33:21 Innovation in High-Stakes Environments39:28 The Future of AI Agents in the Workplace42:25 Cybersecurity Challenges with AI Agents48:18 Innovation Q&AConnect with David* Website: https://www.axonis.ai/ * LinkedIn: https://www.linkedin.com/in/dr-bauer/ Connect with Vit* Substuck: https://anhourofinnovation.substack.com/* LinkedIn: https://www.linkedin.com/in/vit-lyoshin/ * X: https://x.com/vitlyoshin Support the PodcastIf you enjoy the podcast, you can support it by exploring the tools below.These are affiliate links, meaning the show earns a small commission at no extra cost to you.Kit: Email marketing that automates your growth. https://partners.kit.com/xqcc7ca2ahlmGoogle Workspace: Collaborative way of working in the cloud, from anywhere, on any device.https://referworkspace.app.goo.gl/A7wHMeetGeek: Record, transcribe, summarize, and share insights from every meeting. https://get.meetgeek.ai/yjteozr4m6ln For inquiries about sponsoring An Hour of Innovation, email [email protected]

Ask this episode

“What did the council decide on item 42?” — answered with the speaker and the moment.

Transcript

Related