IT and Martech Leaders Become AI Referees When Managing Agents Across 500 Enterprise Applications

In this video you'll learn why enterprise IT leaders are becoming "AI referees" as they face the challenge of managing AI agents across hundreds of applications. Rich explains the three vectors of agent capabilities emerging in the market and why iPaaS-backed agents offer unique advantages for governance and control. Discover why companies with 400-500 applications need centralized orchestration rather than point solutions, and how this fundamental shift is forcing every software domain to reinvent itself in the age of AI.

#EnterpriseAI #AIGovernance #iPaaS #AIOrchestration #AIAgents #ITLeadership #SoftwareIntegration #AIStrategy #ApplicationManagement #AIReferee

Full episode: https://humansofmartech.com/2025/03/25/162-rich-waldron-how-to-build-and-manage-ai-agents/

The collision of AI agents and enterprise software stacks creates messy, complex challenges for tech leaders right now. Rich believes software vendors face a brutal reckoning – you either own critical data or deliver an essential service within a specific domain, or you’re toast. Look around any major department and you’ll see the pattern: Salesforce dominates CRM data, Workday owns the employee record, NetSuite commands financial transactions.

Remember when your CIO discovered your company had somehow accumulated 500+ applications? That nightmare triggered the iPaaS revolution – connect everything without losing your mind. But now we face something potentially worse: AI agent chaos.

“If you’re the marketing ops leader, and your 30 vendors are telling you to go turn on the AI for their application, you sort of become like an AI referee. You have to figure out like, well, do I turn it on for this one and not for this one? And is this one gonna overwrite what occurs here?”

You feel it already. Your inbox floods with vendors begging you to “just turn on our AI capability” – thirty different platforms all promising transformation. Suddenly you’re an unwilling AI referee asking impossible questions:

Which AI systems should I activate first?
What happens when one AI’s decisions contradict another’s?
Are all these systems feeding sensitive data into the same models?
How do I stop someone from accidentally sharing customer SSNs with the wrong AI?
Rich sees three distinct paths forming in this chaotic landscape:

Point-solution AI tools (think: AI-powered SDR applications or marketing campaign optimizers). These smaller vendors need extraordinarily compelling value propositions to survive the coming consolidation.
Enterprise platform AI solutions from major vendors. These work brilliantly within their own domains but struggle with anything outside their walls. You depend entirely on their roadmap priorities.
iPaaS-backed agents that naturally integrate with everything. These provide centralized governance, unified control points, and consistent execution environments.
For IT leaders, the third option solves a critical problem: your marketing ops team can build agents within an established governance framework rather than juggling dozens of disconnected AI systems with different rules and capabilities.

The technical hurdles remain substantial. Every AI agent requires robust integrations capable of executing complex actions. While AI will eventually build these connections independently, we’re years away from that reality. Your organization faces painful experimentation and stack restructuring. Can you accomplish your goals with a single platform? Or are you sliding back into managing a fragmented mess of disconnected systems?

Every software domain must reinvent itself for this new reality. For iPaaS providers like Tray, merely connecting systems no longer suffices – they must support sophisticated AI workloads or perish. The entire software industry scrambles to find stable ground as the landscape transforms beneath their feet.

Видео IT and Martech Leaders Become AI Referees When Managing Agents Across 500 Enterprise Applications канала Humans of Martech Podcast